EA040773B1 - ANTI-TIGIT ANTIBODIES, ANTI-PVRIG ANTIBODIES AND THEIR COMBINATIONS - Google Patents

Description

I. Related Applications

This application claims priority over U.S. Patent Application No. 62/376334, filed Aug. 17, 2016, U.S. Patent Application No. 62/513,771, filed June 1, 2017, U.S. Patent Application No. 62/376,335, filed August 17, 2016 , U.S. Patent Application No. 62/417,217, filed November 3, 2016, U.S. Patent Application No. 62/513,775, filed June 1, 2017, U.S. Patent Application No. 62/477,974, filed March 28, 2017, Application U.S. Patent No. 62/513,916, filed June 1, 2017, and U.S. Patent Application No. 62/538,561, filed July 28, 2017, all of which are incorporated herein by reference in their entirety.

Sequence listing

This application includes a sequence listing that has been submitted electronically in ASCII format and is incorporated herein by reference in its entirety. The specified ASCII copy, created on August 17, 2017, is named 114386-5008-WO_SL.txt and has a size of 590436 bytes.

II. State of the art

Naive T cells must receive two independent signals from antigen presenting cells (APCs) to become productively activated. The first, signal 1, is antigen specific and appears when T-cell antigen recognition receptors encounter the appropriate antigen-MHC complex on the APC. The immune response pathway is determined by a second antigen-independent signal (signal 2) that is delivered via a costimulatory T cell molecule that interacts with its APC-expressed ligand. This second signal can be either stimulatory (positive costimulation) or inhibitory (negative costimulation or co-inhibition). In the absence of a co-stimulatory signal, or in the presence of a co-inhibitory signal, T cell activation is impaired or interrupted, which can lead to a state of antigen-specific non-responsiveness (known as T-cell anergy) or can lead to T-cell apoptotic death.

Pairs of co-stimulatory molecules typically consist of ligands expressed on APCs and their cognate receptors expressed on T cells. The prototypical ligand-receptor pairs of costimulatory molecules are B7/CD28 and CD40/CD40L. The B7 family consists of structurally related cell surface protein ligands that can provide stimulatory or inhibitory contributions to the immune response. Members of the B7 family are structurally linked to an extracellular domain containing at least one immunoglobulin variable or constant domain.

Both positive and negative co-stimulatory signals play a critical role in the regulation of cellular immune responses, and molecules that mediate these signals have proven to be effective targets for immunomodulation. Based on this knowledge, several therapeutic approaches have been developed that involve targeting co-stimulatory molecules and have been shown to be useful in the prevention and treatment of cancer by turning on or off immune responses in cancer patients, and in the prevention and treatment of autoimmune diseases and inflammatory diseases. diseases, as well as rejection in allogeneic transplantation, each of which turns off uncontrolled immune responses or by inducing an off signal by negative costimulation (or co-inhibition) in subjects with these pathological conditions.

Manipulation of signals produced by B7 ligands has shown potential in the treatment of autoimmune diseases, inflammatory diseases, and transplant rejection. Therapeutic strategies include blocking costimulation using monoclonal antibodies to the ligand or costimulatory pair receptor, or using soluble fusion proteins consisting of a costimulatory receptor that can bind and block the corresponding ligand. Another approach is to induce co-inhibition using a soluble inhibitory ligand fusion protein. These approaches rely at least in part on the possible removal of auto- or alloreactive T cells (which are responsible for pathogenic processes in autoimmune disease or transplantation, respectively), presumably because in the absence of costimulation (which induces cell survival genes) T -cells become very susceptible to the induction of apoptosis. Thus, novel agents capable of modulating co-stimulatory signals without compromising the ability of the immune system to defend against pathogens are highly beneficial in the treatment and prevention of such pathological conditions.

Costimulatory pathways play an important role in tumor development. Interestingly, tumors have been shown to elude immune lysis by preventing T cell activation by inhibiting co-stimulatory factors in the B7-CD28 and TNF families, and by recruiting regulatory T cells that inhibit antitumor T cell responses (see Wang ( 2006), Immune Suppression by Tumor Specific CD4 ⁺ Regulatory T cells in Cancer, Semin. Cancer. Biol. 16:73-79; Greenwald, et al. (2005), The B7 Family Revisited, Ann. Rev. Immunol. 23: 515-48, Watts (2005), TNF/TNFR Family Members in Co-stimulation of T Cell Responses, Ann. IRev. Immunol. 23:23-68, Sadum, et al., (2007) Immune Signatures of Murine and Human Cancers Reveal Unique Mechanisms of Tumor Escape and New Targets for Cancer Immunotherapy, Clin Canc Res 13(13): 4016-4025). Such tumor expressed braids

- 1 040773 stimulating molecules have become attractive cancer biomarkers and can serve as tumor-associated antigens (TAA). In addition, costimulatory pathways have been identified as immunological checkpoints that attenuate T-cell dependent immune responses both at the level of initiation and at the level of effector function in tumor metastases. As developed cancer vaccines continue to improve, it is becoming clear that such immunological checkpoints are a major impediment to the ability of vaccines to induce therapeutic antitumor responses. In this regard, co-stimulatory molecules can serve as adjuvants for active (vaccination) and passive (antibody-mediated) cancer immunotherapy, providing strategies to prevent immune tolerance and stimulate the immune system.

Over the past decade, agonists and/or antagonists of various costimulatory proteins have been developed for the treatment of autoimmune diseases, transplant rejection, allergies, and cancer. For example, CTLA4-Ig (abatacept, Orencia®) is approved for the treatment of RA (rheumatoid arthritis), mutated CTLA4-Ig (belatacept, Nulojix®) to prevent acute kidney transplant rejection, and with an anti-CTLA4 antibody (ipilimumab, Yervoy®), recently approved for the treatment of melanoma. Other cost-stimulation regulators have been approved, such as anti-PD-1 antibodies Merck (Keytruda®) and BMS (Opdivo®), have been approved for cancer treatment and are in testing for viral infections.

However, although anti-checkpoint inhibitor monotherapy has been proven, a number of studies (Ahmadzadeh et al., Blood 114:1537 (2009), Matsuzaki et al., PNAS 107(17): 7875-7880 (2010) , Fourcade et al., Cancer Res. 72(4):887-896 (2012) and Gros et al., J. Clinical Invest.124(5):2246 (2014), who studied tumor infiltrating lymphocytes (TIL), demonstrated that TILs typically express multiple checkpoint receptors.Moreover, it is likely that TILs that express multiple checkpoints are in fact the most tumor-reactive.In contrast, non-tumor reactive T cells in the periphery are more likely to express a single checkpoint Checkpoint blockade using monospecific full-length antibodies appears to be non-discriminatory in terms of derepression of tumor-reactive TILs, compared to autoantigen-reactive single expressing T cells, which are thought to be i, contribute to autoimmune toxicity.

One target of interest is PVRIG. PVRIG, also referred to as the poliovirus receptor-associated immunoglobulin domain-containing protein, Q6DKI7 or C7orf15, is a 326 amino acid transmembrane domain protein with signal peptide (ranging from amino acid 1 to 40), extracellular domain (ranging from amino acid 41 to 171), transmembrane a domain (ranging from amino acids 172 to 190); and a cytoplasmic domain (ranging from amino acids 191 to 326). PVRIG binds to poliovirus receptor-related protein 2 (PVLR2, also known as nectin-2, CD112 or herpes virus entry mediator B (HVEB), human cytoplasmic membrane glycoprotein), a PVRIG binding partner.

Another target of interest is TIGIT. TIGIT is a co-inhibitory receptor that is highly expressed on effector and regulatory (Treg) CD4 + T cells, effector CD8 + T cells and NK cells. TIGIT has been shown to attenuate the immune response by (1) direct signaling, (2) induction of ligand signaling, and (3) competition with impaired signaling by the co-stimulatory receptor CD226 (also known as DNAM-1). TIGIT signaling has been best studied in NK cells, where interaction with its cognate ligand, the polio virus receptor (PVR, also known as CD155), has been shown to directly suppress NK cell cytotoxicity via their ITIM (immunoreceptor tyrosine inhibitory motif) cytoplasmic domain. ). TIGIT gene knockout or antibody blockade of the TIGIT/PVR interaction has been shown to increase NK cell death in vitro and exacerbate autoimmune diseases in vivo. In addition to direct effects on T and NK cells, TIGIT can induce PVR-mediated signaling in dendritic or tumor cells, resulting in increased production of anti-inflammatory cytokines such as IL-10. In T cells, TIGIT can also inhibit lymphocyte responses by interfering with the homodimerization of the costimulatory CD226 receptor and competing with it for PVR binding.

TIGIT is expressed at a high level on lymphocytes, including tumor infiltrating lymphocytes (TIL) and Treg, which infiltrate various types of tumors. PVR is also widely expressed in tumors, indicating that the TIGIT-PVR signaling axis may be a major immunosurveillance evasion mechanism for cancer. Notably, the expression of TIGIT is closely correlated with the expression of another important co-inhibitory receptor, PD1. TIGIT and PD1 are co-expressed on the TIL of numerous human and mouse tumors. In contrast to TIGIT and CTLA4 PD1, inhibition of T cell responses is not associated with competition for ligand binding to the costimulatory receptor.

Accordingly, TIGIT is an attractive target for monoclonal antibody therapy.

- 2 040773 lamy and, moreover, in combination with additional antibodies, including anti-PVRIG antibodies.

III. Brief description of the essence of the invention

Accordingly, in one aspect, the invention provides compositions comprising an antigen binding domain that binds to human TIGIT (SEQ ID NO: 97) comprising a heavy chain variable domain comprising SEQ ID NO: 160 and a light chain variable domain comprising SEQ ID NO: : 165. In addition, the antigen-binding domain contains a heavy chain variable domain containing SEQ ID NO: 150, and a light chain variable domain containing SEQ ID NO: 155. In addition, the antigen-binding domain contains a heavy chain variable domain containing SEQ ID NO: 560, and a light chain variable domain containing SEQ ID NO: 565.

In the following aspect, the invention relates to a composition containing antibodies containing a heavy chain containing VH-CH1-hinge-CH2-CH3, and the specified VH contains SEQ ID NO: 160, and a light chain containing VL-VC, and the specified VL containing SEQ ID NO: 165 and VC is either kappa or lambda. In addition, the antibody may contain a heavy chain containing VH-CH1 hinge-CH2-CH3, and VH contains SEQ ID NO: 150; and a light chain containing a VL-VC, said VL having SEQ ID NO: 159 and the VC being either kappa or lambda. In addition, the antibody may contain a heavy chain containing VH-CH1-hinge-CH2-CH3, and VH contains SEQ ID NO: 560; and a light chain containing a VL-VC, said VL having SEQ ID NO: 565 and the VC being either kappa or lambda.

In some aspects, the sequence CH1-hinge-CH2-CH3 is selected from human IgGl, IgG2 and IgG4 and variants thereof. In some aspects, the heavy chain is SEQ ID NO: 164 and the light chain is SEQ ID NO: 169.

In a further aspect, the compositions may further comprise a second antibody that binds to a human checkpoint receptor protein, which may be human PD-1 or human PVRIG. The second antibody may comprise an antigen binding domain comprising a heavy chain variable domain having SEQ ID NO: 5 and a light chain variable domain having SEQ ID NO: 10, or a heavy chain having SEQ ID NO: 9 and a light chain having SEQ ID NO: 14.

In a further aspect, the invention relates to nucleic acid compositions comprising a first nucleic acid encoding a heavy variable domain comprising SEQ ID NO: 160 and a second nucleic acid encoding a light chain variable domain comprising SEQ ID NO: 165. Alternatively, nucleic acid compositions contain a first nucleic acid encoding a heavy chain variable domain comprising SEQ ID NO: 150 and a second nucleic acid encoding a light chain variable domain comprising SEQ ID NO: 155. Alternatively, the nucleic acid compositions comprise a first nucleic acid encoding a heavy chain variable domain chain containing SEQ ID NO: 560 and a second nucleic acid encoding a light chain variable domain containing SEQ ID NO: 565.

In a further aspect, the invention relates to expression vector compositions containing nucleic acid composition data, as well as, for example, a first expression vector containing a first nucleic acid and a second expression vector containing a second nucleic acid, or alternatively an expression vector that contains both first and second nucleic acids.

In an additional aspect, the invention relates to host cells containing expression vector compositions, and methods for producing antibodies, including culturing host cells under conditions in which antibodies are produced, and isolating the antibody.

In a further aspect, the invention relates to anti-PVRIG antibodies comprising a heavy chain having SEQ ID NO: 9 and a light chain having SEQ ID NO: 14. The invention also relates to antibodies having a heavy chain having SEQ ID NO: 19; and a light chain having SEQ ID NO: 24.

In a further aspect, the anti-PVRIG antibody (either CHA.7.518.1.H4 (S241P) or CHA.7.538.1.2.H4 (S241P) is co-administered with a second antibody that binds to a human checkpoint receptor protein, such as an antibody which binds PD-l.

In a further aspect, an anti-PVRIG antibody (either CHA.7.518.1.H4 (S241P) or CHA.7.538.1.2.H4 (S241P)) is co-administered with a second antibody that binds to a human checkpoint receptor protein, such as an antibody , which binds the human TIGIT, such as CPA.9.086 or CPA.9.083, or CHA.9.547.13.

In a further aspect, the invention relates to nucleic acid compositions comprising a first nucleic acid encoding a heavy chain of either CHA.7.518.1.H4 (S241P) or CHA.7.538.1.2.H4 (S241P)), and a second nucleic acid encoding a light chain. chain or CHA7.518.1.H4 (S241P), or CHA7.538.1.2.H4 (S241P), respectively.

In a further aspect, the invention relates to expression vector compositions containing nucleic acid composition data, as well as, for example, a first expression vector containing a first nucleic acid and a second expression vector containing a second nucleic acid.

- 3 040773 acid, or, alternatively, an expression vector that contains both the first and second nucleic acids.

In an additional aspect, the invention relates to host cells containing expression vector compositions, and methods for producing antibodies, including culturing host cells under conditions in which antibodies are produced, and isolating the antibody.

In another aspect, the invention relates to methods comprising: a) providing a population of cells from a tumor sample in a patient; b) staining said population with labeled antibodies that bind: i) TIGIT protein; ii) PVR protein; iii) PD-1 protein; iv) PD-L1 protein; and v) isotype control; c) performing a fluorescence activated cell sorting (FACS) assay; d) for each of TIGIT, PVR, PD-1 and PD-L1, determining the percentage of cells in said population that express the protein relative to said isotype control antibody; moreover, if the percentage of positive cells is> 1% for all 4 receptors, e) the introduction of antibodies to TIGIT and PD-1 to the specified patient.

In another aspect, the invention relates to methods comprising: a) providing a population of cells from a tumor sample in a patient; b) staining said population with labeled antibodies that bind: i) PVRIG protein; ii) PVRL2 protein; iii) PD-1 protein; iv) PD-L1 protein; and v) isotype control; c) fluorescence activated cell sorting (FACS) assay method; d) for each of PVRIG, PVRL2, PD-1 and PD-L1, determining the percentage of cells in said population that express the protein relative to said isotype control antibody; moreover, if the percentage of positive cells is >1% for all 4 receptors, e) the introduction of antibodies to PVRIG and PD-1 to the specified patient.

In another aspect, the invention relates to methods comprising: a) providing a population of cells from a tumor sample in a patient; b) staining said population with labeled antibodies that bind: i) PVRIG protein; ii) PVRL2 protein; iii) TIGIT protein; iv) PVR protein; and v) isotype control; c) performing analysis using flow cytometry (FACS); d) for each of PVRIG, PVRL2, TIGIT, and PVR, determining the percentage of cells in said population that express the protein relative to said isotype control antibody; moreover, if the percentage of positive cells is > 1% for all 4 receptors, e) the introduction of antibodies to PVRIG and TIGIT specified patient.

IV. Brief description of graphic materials

In FIG. 1 depicts the full length human PVRIG sequence (showing two different methionine start points). The signal peptide is underlined, the EVA is underlined twice.

In FIG. 2 depicts the sequence of human poliomyelitis virus receptor-associated protein 2 (PVLR2, also known as nectin-2, CD112 or herpes virus entry mediator B (HVEB)), a binding partner for PVRIG. PVLR2 is a human cytoplasmic membrane glycoprotein.

In FIG. 3A and B show the variable heavy and light chains and the sequences of vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2 and vlCDR3 of each of the listed CHA antibodies of the invention, CHA.7.518.1.H4(S241P), and CHA7.538.1.2. H4(S241P).

In FIG. 4A and B, anti-PVRIG antibodies increase T cell proliferation in MCL (mixed lymphocyte culture). The percentage of cells with low levels of CFSE (carboxyfluorescein succinimidyl ether) is depicted in the SCL reaction treated with the indicated antibodies to PVRIG. Each graph represents one individual CD3 T cell donor. The experiments are described in Example 23 USSN 15/048967, incorporated herein by reference.

In FIG. 5 A and B binding characteristics of anti-PVRIG hybridoma antibodies to modified HEK hPVRIG cell lines, original HEK cells, and Jurkat cells. HEK OE denotes HEK hPVRIG cells, HEK par denotes original HEK cells. For Jurkat data, gMFIr indicates the fold difference in the geometric CIF of anti-PVRIG antibody staining relative to their controls. The concentration indicates at which gMFIr was calculated. Lack of binding indicates that the antibody does not bind to the cell line under investigation. The isolated antibodies are the top four antibodies of interest.

Fig. 6A and B. Anti-PVRIG hybridoma antibody binding characteristics to primary human PBMC, cynomolgus monkey overexpressing cells, and primary cynomolgus PBMC. Expi cyno OE refers to expi cells transiently transfected with cPVRIG, expi par refers to original expi cells. gMFIr indicates the fold difference in the geometric CIF of anti-PVRIG antibody staining relative to their controls. Concentrations indicate at which gMFIr was calculated. He tested indicates antibodies that were not tested due to lack of binding to human HEK hKVRIG, expi cPVRIG cells, or did not meet the requirements for binding to PBMC subpopulations. The isolated antibodies are the top four antibodies of interest. The experiments are described in Example 21 USSN 15/048967, incorporated herein by reference.

- 4 040773

Fig. 7A and B. Summary of PVRIG antibody blocking ability in a competitive flow cytometric assay. Shown IR ₅₀ inhibition. No _IC50 indicates that these antibodies are non-blocking. The isolated antibodies are the top four antibodies of interest. The experiments are described in Example 21 USSN 15/048967 incorporated herein by reference.

In FIG. 8A and B TIL were co-cultured with 624 melanoma cells at 1:1 E:T for 18 h in the presence of anti-PVRIG Ab (CPA.7.021; 10 μg/ml) anti-TIGIT (10A7 clone, 10 μg/ml) or in combination. The supernatant was harvested and tested in a Th1/Th2/Th17 Cytometric Multiplex Assay (Cytometric Bead Array) to determine secreted cytokines. The levels of IFN-γ (A) and TNF (B) were measured. Treatments were compared using Student's t-test (*P<0.05, **P<0.01) of three samples.

In FIG. 9 A to F TIL MART-1 or 209 were co-cultured with 624 melanoma cells at 1:1 E:T for 18 h in the presence of anti-PVRIG Ab (CPA.7.021, 10 μg/ml), anti-DNAM1 ( clone DX11, BD Biosciences cat # 559787; 10 µg/mL) or in combination. The supernatant was harvested and tested in a Th1/Th2/Th17 Cytometric Multiplex Assay (Cytometric Bead Array) to determine secreted cytokines. The levels of IFN-γ (A, D) and TNF (B, E) were measured. TIL was stained for surface expression of CD 137 (C, F).

In FIG. 10A and B TIL (F4) were co-cultured with 624 melanoma cells at 1:3 E:T for 18 h in the presence of anti-PVRIG Ab (CPA.7.021, 10 μg/ml) anti-TIGIT (clone 10A7, 10 μg /ml), anti-PD1 (mAb 1B8, Merck, 10 µg/ml) or in combination. The supernatant was harvested and tested in a Th1/Th2/Th17 Cytometric Multiplex Assay (Cytometric Bead Array) to determine secreted cytokines. The levels of IFN-γ (A) and TNF (B) were measured.

In FIG. 11 A to E depict four humanized sequences for each of CHA.7.518, CHA.7.524, CHA.7.530, CHA7.538_1, and CHA7.538_2. All humanized antibodies contain an H4(S241P) substitution. It should be noted that the light chain for CHA.7.538_2 is the same as for CHA.7.538_1. The H1 of each is a CDR replacement with no changes in the human framework. Subsequent sequences defining changes in the framework region are shown in bold. The CDR sequences are shown in bold. CDR definitions are ALM from www.bioinf.org.uk/abs/. Human germline and fusion sequences from IMGT®, the international information system ImMunoGeneTics® www.imgt.org (founder and director: Marie-Paoulet Lefranc, Montpellier, France). Residue numbering is shown as sequential (seq) or according to Chothia from www.bioinf.org.uk/abs/ (ALM). b denotes a hidden side chain; p denotes partially hidden; i denotes the side chain at the interface between the VH and VL domains. Sequence differences between human and mouse germlines are marked with an asterisk (*). Potential additional mutations in the structures are noted below the sequence. Potential changes in the CDR sequences noted below each CDR sequence as noted in the figure (# deamidation substitutions: Q/S/A, this may prevent asparagine (N) deamidation. @ tryptophan oxidation substitutions: Y/F/H; they may prevent tryptophan oxidation; @ methionine oxidation substitutions: L/F/A).

In FIG. 12 A to E shows the alignment of the humanized sequences of three CHA antibodies: CHA.7.518, CHA.7.538.1 and CHA.7.538.2.

In FIG. 13 shows schemes for combining humanized VH and VL anti-CHA antibodies. ChimVH and chimVL are mouse heavy and light variable sequences fused to a human IgG constant region.

Fig. 14. Characterization of binding of anti-PVRIG hybridoma antibodies to human primary PBMCs, cynomolgus monkey overexpressing cells, and primary cynomolgus monkey PBMCs. Expi cyno OE refers to expi cells transiently transfected with cPVRIG, expi par refers to original expi cells. gMFIr indicates the fold difference in the geometric CIF of PVRIG antibody staining relative to their controls. The concentrations indicate at which gMFIr was calculated. Not tested indicates antibodies that were not tested because they did not bind to human hKVRIG Hek, expi cPVRIG cells, or did not meet the requirements for binding to PBMC subpopulations. The isolated antibodies represent the four antibodies that were humanized (see FIG. 24). The experiments are described in Example 21 USSN 15/048967 incorporated herein by reference.

Fig. 15. Summary of blocking ability of PVRIG antibodies in a competition analysis based on flow cytometry. The IC50 of inhibition is shown. None of the IC50s indicates that these antibodies are non-blocking. The isolated antibodies represent four antibodies for which humanization was performed (see Fig. 24).

Fig. 16. Pooled activities of selected anti-PVRIG antibodies in NK cell cytotoxicity assays against Reh and MOLM-13 cells. The fold change in cytotoxicity compared to control was calculated by dividing the absolute death rate (%) using an anti-PVRIG antibody by the absolute

- 5 040773 death rate (%) using control antibody. The fold change is calculated based on a 5:1 effector-target ratio.

Fig. 17. Sequence alignment of PVRIG orthologs. Aligned sequences of the extracellular domain of human PVRIG, cynomolgus monkey, marmoset and rhesus monkey. Differences between human and cynomolgus monkey sequences are highlighted in yellow.

Fig. 18. Binding of anti-human PVRIG antibodies to cynomolgus, human, cynomolgus/human PVRIG variants. Antibody binding to wild-type cynomolgus monkey PVRIG (·), H61R cynomolgus monkey PVRIG (), P67S cynomolgus monkey PVRIG ( ^A ), L95R/T97I cyno PVRIG (▼), and wild-type human PVRIG (♦) is shown. ELISA signals are displayed depending on the concentration of antibodies.

Fig. 19. Correlation of epitope group and cross-reactivity of antibodies against human PVRIG with cynomolgus macaque.

Fig. 20 A to B. (A) Specificity of CHA.7.518.1.H4(S241P) for Hek cells modified to overexpress PVRIG and native HEK cells. The data show the absolute geometric values of SIF (gSIF) as a function of the increase in antibody concentration. (B) Specificity of CHA.7.538.1.2.H4 (S241P) for HEK cells modified to overexpress PVRIG and native HEK cells. The data show the absolute geometric values of SIF (gSIF) as a function of the increase in antibody concentration.

In FIG. 21 A-B depict the ability of CHA.7.518.1.H4 (S241P) (A) and CHA.7.538.1.2.H4 (S241P) (B) to bind Jurkat cells that endogenously express PVRIG confirmed by RNA expression. (A) Binding of CHA.7.518.1.H4 (S241P) to Jurkat cells. The data show the absolute geometric values of SIF (gSIF) as a function of the increase in antibody concentration. Isotype staining is depicted as a negative control. (B) Binding of CHA.7.538.1.2.H4 (S241P) to Jurkat cells. The data show the absolute geometric values of SIF (gSIF) as a function of the increase in antibody concentration. Isotype staining is depicted as a negative control. Both antibodies are able to bind Jurkat cells with comparable affinity to HEK hPVRIG cells.

In FIG. 22 shows the ability of CHA.7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) to bind CD8 T cells that have been quantified by exposure to the CMV peptide (494-503, NLVPMVATV), and which endogenously express PVRIG confirmed by RNA expression. Binding of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) to CMV peptide-expanded CD8 T cells. The data show the absolute geometric values of CIF (rCIF) as a function of increasing antibody concentration. Isotype staining is depicted as a negative control.

Fig. 23A-B. (A) Specificity of CHA.7.518.1.H4 (S241P) for expi cells modified to overexpress cynomolgus monkey PVRIG and parental expi cells. The data show the absolute geometric values of SIF (gSIF) as a function of the increase in antibody concentration. Specificity of CHA.7.538.1.2.H4 (S241P) for expi cells modified to overexpress cynomolgus monkey PVRIG and parental expi cells. The data show measurements of the absolute geometric values of CIF (gCIF) as a function of increasing antibody concentration.

Fig. 24A-B. (A) Blocking of PVRIG Fc in HEK cells by CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P). The data show the percentage of PVRIG Fc with HEK cells as a function of the increase in antibody concentration compared to the maximum signal induced by PVRIG Fc and secondary background alone. (B) Effect of CHA.7.544 on PVRIG Fc binding to HEK cells. The data demonstrate the absolute gSIF obtained from the binding of PVRIG Fc to HEK cells at increasing concentrations of CHA.7.544. The amount of PVRIG Fc binding was detected using anti-mouse Fc secondarily conjugated to Alexa 647.

Fig. 25A-B. (A) Blocking of PVRL2 Fc with HEK hPVRIG cells with CHA7.518.1.H4 (S241P), CHA7.538.1.2.H4 (S241P) and CHA.7.530.3. The data show the percentage binding of PVRL2 Fc to HEK hPVRIG cells as a function of the increase in antibody concentration compared to the maximum signal induced by PVRL2 Fc and secondary background alone. (B) Effect of CHA.7.544 on PVRL2 Fc binding to HEK hPVRIG cells. The data show the percentage binding of PVRL2 Fc to HEK hPVRIG cells as a function of the increase in antibody concentration compared to the maximum signal induced by PVRL2 Fc and secondary antibody as background.

Fig. 26. Percentage of binding of Alexa 647-conjugated CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) compared to their maximum signal when Jurkat cells were pre-incubated with unconjugated CHA7.518.1.H4 (S241P) is shown. ), CHA7.538.1.2.H4 (S241P) and isotopic control.

Fig. 27A-B. A) Humanized antibodies to PVRIG, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), increase CD4 + T cell proliferation. Representative data (n>2) show the percentage of proliferating CD4+ T cells with low CFSE levels (mean plus standard

- 6 040773 dart) from a single donor of human CD4+ T cells when co-cultivated with CHO-S OKT3 hPVRL2 cells in the presence of anti-DNAM-1 antibody or various anti-PVRIG antibodies or IgG isotype controls. The dotted line indicates the baseline percentage of CD4+ T cells with low CFSE levels proliferating after treatment with the isotype control human IgG4 antibody. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies. (B) Humanized anti-PVRIG antibodies, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), increase CD4+ T cell proliferation in a hPVRL2 dependent manner. Representative data (n>2) show the percentage of proliferating CD4+ T cells with low CFSE (mean plus standard deviation) from a single human CD4+ T cell donor in response to co-culture with CHO-S OKT3 stock or CHO-S OKT3 hPVRL2 cells in the presence of anti-DNAM-1 antibody or various anti-PVRIG antibodies or IgG isotype controls. The dotted line indicates the baseline percentage of CD4+ T cells with low CFSE levels proliferating after treatment with isotype antibodies of either human IgG4 or mouse IgG1. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies.

Fig. 28A-C. (A) Humanized antibodies to PVRIG, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), increase CD8+ T cell proliferation. Representative data (n>2) show the percentage of proliferating CD8+ T cells with low CFSE levels (mean plus standard deviation) from a single donor (donor 232) of human CD8+ T cells when co-cultured with OKT3 hPVRL2 CHO-S cells in the presence of anti -DNAM-1 antibodies or various anti-PVRIG antibodies or IgG isotype controls. The dotted line indicates the baseline percentage of low-CFSE CD8+ T cells proliferating after treatment with mouse IgG1 or human IgG4 isotype antibodies. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies. (B) Humanized antibodies to PVRIG, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), increase CD8 + T cell proliferation. Representative data (n>2) show the percentage of proliferating CD8+ T cells with low CFSE (mean plus standard deviation) from a single donor (donor 234) of human CD8+ T cells when co-cultured with OKT3 hPVRL2 CHO-S cells in the presence of anti -DNAM-1 antibodies or various anti-PVRIG antibodies or IgG isotype controls. The dotted line indicates the baseline percentage of low-CFSE CD8+ T cells proliferating after treatment with mouse IgG1 or human IgG4 isotype antibodies. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies. (C) Humanized anti-PVRIG antibodies, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), increase IFN-γ secretion from CD8+ T cells. Representative data (n>2) show pg/ml IFN-γ produced (mean plus standard deviation) from three different human CD8+ T cell donors (donors 231, 232 and 234) when co-cultured with hPVRL2 OKT3 CHO-S cells in the presence of an anti-DNAM-1 antibody, or various anti-PVRIG antibodies, or IgG isotype controls. The dotted line indicates the initial production of IFNγ after treatment with isotype human IgG4 antibody. The numbers refer to the percentage increase in IFN-γ secretion after treatment with anti-PVRIG antibodies compared to an IgG4 isotype control.

Fig. 29. Humanized anti-PVRIG antibodies, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), consistently increase CD4+ T cell proliferation in multiple donors, while CHA.7.530.3 and CHA.7.544 they don't. Percent proliferation compared to isotype control was calculated by dividing the percentage of CD4+ T cells with low CFSE after anti-PVRIG antibody treatment by the percentage after isotype antibody treatment for each donor. Percent proliferation for isotype antibody treatment was set to zero. Each symbol on the graph represents a different donor.

Fig. 30A-D. (A) Dose-dependent effect of humanized antibodies to PVRIG,

CHA7.518.1.H4(S241P) and ChA7.538.1.2.H4(S241P) on CD4+ T cell proliferation. Representative data (n>2) from two different human donors show the mean percentage of proliferating CD4+ T cells after dose changes from 66 nM to 0.726 nM with human IgG4 isotype, CHA.7.518.1.H4(S241P) or CHA antibodies. 7.538.1.2.H4(S241P). Estimated EC50 is within one digit in the nM range. (B) Dose-dependent effect of humanized anti-PVRIG antibodies, CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P), on CD8+ T cell proliferation. Representative data (n>2) from two different human donors show the average percentage of proliferating CD8+ T cells after dose changes from 66 nM to 0.264 nM with human IgG4 isotype, antibodies CHA.7.518.1.H4(S241P), CHA.7.38. 1.2, or CHA.7.544. Estimated EC50 is within one digit

- 7 040773 ry in the nm range.

Fig. 31A-C. (A) Flow cytometry analysis of TIGIT and PVRIG expression on TIL and PVR, PVRL2 expression on melanoma cell line 624. Values are the ratio of mean fluorescence intensity (MFI) and isotype control. (B-C) Representative experiment demonstrating IFN-γ (B) and TNF (C) secretion by TIL when co-cultured with 624 melanoma cells at 1:3 E:T for 18 h in the presence of isotype control, anti-TIGIT (30 µg/ml) or anti-PVRIG Ab (10 µg/ml) as monotherapy (blue bars) or in combination with anti-TIGIT (green bars). The percentage effect of Ab monotherapy was compared with mIgG1 isotype therapy, and the percentage effect of Ab combination therapy was compared with anti-TIGIT monotherapy.

Fig. 32A-H. TIL (209-gp100/463-F4-gp100) were co-cultured with 624 melanoma cells at 1:3 E:T for 18 h in the presence of anti-PVRIG Ab CHA7.518.1.H4 (S241P) or CHA.7.538 with or without anti-TIGIT (aTIGIT) in combination and tested for cytokine secretion. The percentage effect of Ab therapy was compared with isotype control therapy and the average of 5 experiments (F4) or 6 experiments (209) was plotted. For each treatment, a two-tailed paired Student's t-test was calculated versus isotype or in combination versus anti-TIGIT alone, p-values are indicated.

Fig. 33A-B. (A) Humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) and anti-TIGIT antibody increased CD4+ T cell proliferation compared to treatment with antibody alone. Representative data (n>2) show the percentage of proliferating CD4+ T cells with low CFSE (mean plus standard deviation) from a single donor (donor 143) of human CD3+ T cells when co-cultured with CHO-S OKT3 hPVRL2 cells. The dotted line indicates the baseline percentage of low-CFSE CD4+ T cells proliferating after treatment with the isotype control human IgG4 antibody. (B) Humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P), and anti-TIGIT antibody increased CD4+ T cell proliferation compared to treatment with antibody alone. Representative data (n>2) show the percentage of proliferating CD4+ T cells with low CFSE (mean plus standard deviation) from a single donor (donor 201) of human CD4+ T cells when co-cultured with CHO-S OKT3 hPVRL2 cells. The dotted line indicates the baseline percentage of low-CFSE CD4+ T cells proliferating after treatment with the isotype control human IgG4 antibody. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies.

Fig. 34A-B. (A): The combination of a humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) and an anti-TIGIT antibody increases CD8+ T cell proliferation. Representative data (n>2) show the percentage of proliferating CD8+ T cells with low CFSE (mean plus standard deviation) from a representative donor (donor 232) of human CD8+ T cells when co-cultured with CHO-S OKT3 hPVRL2 cells. The dotted line indicates the baseline percentage of low-CFSE CD8+ T cells proliferating after treatment with the human IgG4 isotype antibody. The numbers refer to the percentage increase or decrease in proliferation resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies. (B) The combination of a humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) and an anti-TIGIT antibody increases IFN-γ secretion from CD8+ T cells. Representative data (n>2) show pg/ml IFN-γ produced (mean plus standard deviation) from a representative human CD8+ T cell donor (donor 232) when co-cultured with CHO-S OKT3 hPVRL2 cells. The dotted line indicates the initial production of IFN-γ after treatment with isotype human IgG4 antibody. The numbers refer to the percentage increase or decrease in IFN-γ secretion resulting from treatment with anti-PVRIG or anti-DNAM-1 antibodies, respectively, compared to the corresponding isotype control antibodies.

In FIG. 35 is a diagram of the experimental system of Example 2(3).

In FIG. 36A-C are histograms depicting expression levels of PVRIG (using human anti-PVRIG CHA.7.538.AF647), TIGIT (using human anti-TIGIT eBioscience cat. 179500-41), and DNAM-1 (using anti-human CD226- APC Cat 338312 Biolegend) at TIL. The fold change in expression is compared to an isotype (Iso) control.

Fig. 37. Summary plot of the effect of anti-PVRIG antibodies on IFN-γ secretion from TIL. TILs were co-cultured with PVRL2 overexpressing CHO-S HLA-A2/B2M cells at an E:T ratio of 1:3 for 18 h in the presence of anti-PVRIG antibodies (c518, c538 and 544) or with anti-TIGIT antibody. Each point represents the mean of IFN-γ secretion data from the same TIL from different experiments. The indicated percentage differs from each antibody treatment compared to the isotype control. For each treatment, a two-tailed paired Student's t-test was calculated versus 544 or in combination versus anti-TIGIT alone, p-values of the

- 8 040773 na. Number of experiments defined in advance per TIL; 209 (N=3), F4 (N=2), F5 (N=3), and MARTI (N=2).

Fig. 38. Generalized plot of the effect of C518 and c538 as a function of dose on TNF-α secretion from TIL. TILs were co-cultured with PVRL2-overexpressing CHO-S HLA-A2/B2M cells at a 1:3 effector-target ratio for 18 h in the presence of anti-PVRIG antibodies (c518, c538 or isotype control) as described in Example 2 (3).

Fig. 39A-C. TILs were co-cultured with target CHO-S HLA-A2/B2M cells overexpressing PVRL2 at an E:T ratio of 1:3 for 18 h in the presence of anti-PVRIG antibodies (c518, c538 and 544) or with anti-TIGIT antibody. . The percentage shown in the tables above is the difference in effect on cytokine secretion from TIL of each antibody treatment compared to its isotype control. The first experiment is shown in Fig. A and B, and the second experiment - in Fig. WITH.

Fig. 40. CHO-S OKT3 co-cultivation assay method. CFSE-labeled CD3+ T cells were co-cultured with CHO-S-OKT3-PVRL2 or mock-transfected cells for 5 days. The effect of anti-PVRIG Abs on T cell proliferation and cytokine secretion was analyzed.

Fig. 41. Effect of anti-PVRIG antibodies on IFN-γ secretion on CHO-OKT3 PVRL2 cells in responder versus non-responder donor. CD3+ cells from 2 different donors were co-cultured with CHO-S-PVRL2 cells at 5:1 E:T for 5 days in the presence of anti-PVRIG Ab and examined for cytokine secretion and T cell proliferation. (A) Donor responding to treatment for which we observed an effect for anti-PVRIG Ab. (B) A non-responder donor for whom we observed no effects for At treatment.

Fig. 42. Effect of anti-PVRIG antibodies on CD4 and CD8 proliferation in a responding donor. CFSE-labeled CD3+ T cells were co-cultured with CHO-S-PVRL2 cells at 5:1 E:T for 5 days in the presence of anti-PVRIG Ab or anti-TIGIT Ab. The effect on proliferation of CD4 or CD8 gated T cells was assessed by flow cytometry. Percentages of proliferating cells (low CFSE) (A) or total cells, (B) CD4+ low CFSE or CD8+ low CFSE are shown.

Fig. 43. The effect of anti-PVRIG antibodies on IFN-γ secretion or CD8 proliferation in a responding donor is shown. CD3+ cells were co-cultured with CHO-SPVRL2 cells at 5:1 E:T for 5 days in the presence of anti-PVRIG Ab and examined for (A) cytokine secretion and (B) T cell proliferation. The percentage effect of Ab therapy was compared to isotype control therapy and the mean of 5 donors responding to treatment (treatment responders) was presented. (C) IFN-γ levels from the same 5 donors when co-cultured with CHOS-OKT3 PVRL2, as described in sections A and B, when treated with isotype compared with anti-PVRIG Ab. The p-value is the ratio of the two-tailed Student's t-test.

In FIG. 44 shows a summary table of the effect of Ab treatment for the studied donors (n = 10). The percentages shown are the treatment effect of Abs for the specific indication (listed in the column headings) compared to the corresponding isotype control, responder donors (donors #3, 72226345 and ES_001), who are considered to be responders for whom some anti-PVRIG Abs (mainly CHA. 7.518) increase IFN-γ or proliferation against isotype control.

In FIG. 45A-B depict the results of experiments with several antibodies. Affinities (nM) are shown in A, HEK hPVRIG cells being HEK cells transformed with hPVRIG as discussed herein and Jurkat cells expressing endogenous hPVRIG. (B) shows hSIF using 4 different antibodies against donor 1 primary CD8 T cells and (C), donor 2 primary CD8 T cells.

In FIG. 46A-B depict TIGIT interactions with CHO cells. (A) Human TIGIT Fc protein binds to CHO cells. Differentiated concentrations of human TIGIT Fc and IgG1 synagis control were assessed by their ability to bind to CHO cells in a FACS-based binding assay. (B) Human PVR is expressed on activated CD4 T cells. CD4 T cells were co-cultured with CHO cells expressing OKT3 scFv antibodies and activated for 5 days. On day 5, CD4 T cells were analyzed for PVR expression and CFSE dilutions.

In FIG. 47A-C depict anti-tumor anti-mPVRIg and anti-PDL-1 antibody responses in the CT26 tumor model. A-V. Groups of 10 BALB/c mice were injected subcutaneously with 5x105 CT26 cells. After tumors were measured on day 4, mice were randomized (mean tumor volume per group 40 mm3) and then treated with the prescribed mAb (100 or 200 µg/dose, i/p) followed by additional doses at 7, 11 , 14, 18 and 21 days. A. Groups were treated with 6 doses of individual agents. AntiPDL-1 versus control ***p<0.0001. Tumor volumes are presented as mean volume + SOS. B. Tumor volumes were measured twice a week. The number of tumor-free mice (TF) per group is reported. C. Survival rates of designated groups; anti-PDL1 versus con

- 9 040773 troll **p = 0.005.

In FIG. 48A-C depict anti-tumor responses of a combination of anti-PVRIG and anti-PDLl antibodies in the CT26 tumor model. A-V. Groups of 10 BALB/c mice were injected subcutaneously with 5x105 CT26 cells. After tumors were measured on day 7, mice were randomized (mean tumor volume per group - 75 mm ³ ) and then treated with the prescribed mAb (300 µg/dose, i/p) followed by additional doses at 11, 14 , 18, 21 and 25 days. A. Groups were treated with 6 doses of combination agents. Ahtu-PDL-1+ mAb 407 compared with control, p = 0.0005; anti-PDL-1 and mAb 406 compared with control, p = 0.056. C. Tumor volumes were measured x3 per week. The number of tumor-free mice (TF) per group is reported. C. survival rates of certain groups; antiPDL-1 + mAb 407 vs control * p = 0.0088.

In FIG. 49A-D depict the amino acid sequences and nucleic acid sequence for the heavy chain variable region (A and B, respectively) and the amino acid sequences and nucleic acid sequence for the light chain variable region (C and D, respectively) for AB-407 (BOJ-5G4- F4).

In FIG. 50 depicts the amino acid sequences of human IgG1 constant domains (with some useful amino acid substitutions), IgG2, IgG3, IgG4, IgG4 with a hinge variant that finds particular use in the present invention, and kappa and lambda light chain constant regions.

In FIG. 51 shows the protein sequences of human and cynomolgus macaque TIGIT EVA (referred to as supo) and human PVR EVA.

Fig. 52. A summary table of bindings from flow cytometry for anti-TIGIT fab is shown. All unique ELISA-positive fab were analyzed by flow cytometry. Mean fluorescence intensity (MFI) was measured for Expi293 cells overexpressing human or cynomolgus TIGIT, as well as parental Expi293 cells. The SIF coefficient for target-specific and non-specific binding was calculated. Data for selected clones are shown.

In FIG. 53A and B depict the sequences of anti-TIGIT antibodies. Unless otherwise noted, CDRs use IMGT numbering (including antibodies from the sequence listing.

Fig. 54. FACS KD results of anti-TIGIT mAb binding to Expi293 cells overexpressing human TIGIT as described in Example 12 are shown.

Fig. 55. FACS KD results of mAb binding to cynomolgus monkey Expi293 cells overexpressing TIGIT are shown.

Fig. 56. Shown are the results from Example 14, which show the obtained kinetic rate constants and equilibrium dissociation constants, where the data were sufficiently reliable to estimate the binding constants.

In FIG. 57A-B depict the results of binding of the human PVR-Fc variant to Expi293 cells overexpressing human TIGIT in Example 4. FIG. A (left): Binding curve for human PVR-m2aFc construct titrated with human TIGT-overexpressing Expi293 cells. KD and 95% confidence interval are shown. Fig. B (right): Binding curve for human PVR-h1Fc construct titrated with human TIGT-overexpressing Expi293 cells. KD and 95% confidence interval are shown.

Fig. 58. Table of phage antibodies inhibiting binding of human PVR-m2aFc to human TIGITs overexpressed on Expi293 cells is shown. The mAbs were tested against a known blocking reference antibody (BM26) as well as an IgG4 isotype control antibody (Synagis). Yes means that the mAb inhibited hPVR similarly to BM26.

Fig. 59. Shown is a table of IC50 values of anti-TIGIT hybridoma antibodies inhibiting the binding of human PVR-hlFc to human TIGIT overexpressed on Expi293 cells. Values are representative of one of two independent experiments. The IC50 results for two independently conducted experiments differ in the range of only 1.2-2 times.

Fig. 60. The results of Example 6 are shown showing that phage and BM derived anti-human TIGIT antibodies CPA.9.027, CPA.9.049, CPA.9.059, BM26 and BM29 increase IL-2 signaling. BM26 and BM29 belong to the human IgG4 isotype (hIgG4 with S241P variant). Representative data (n>2) demonstrate the ESE (mean +/- standard deviation) of the luciferase signal from a 6-hour co-culture of human Jurkat IL-2-RE human TIGIT luciferase cells and human Cho-Κι PVR aARC cells. The concentration of each antibody was 10 μg/ml.

Fig. 61. Additional results of Example 6 are shown showing that phage and BM hIgG4-derived anti-human TIGIT antibodies CPA.9.027, CPA.9.049, CPA.9.059, BM26 and BM29 increase IL-2 signaling in a dose-dependent manner. BM26 and BM29 belong to the hIgG4 isotype. Representative data (n>2) demonstrate the ESE (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of human Jurkat IL-2-RE luciferase cells and human aAPC CHO-K1 PVR cells. For each antibody, 10-point, 2-fold dilution series were used, starting at 20 μg/ml.

- 10 040773

Fig. 62. The results of Example 6 are shown, showing that hybridoma and BM-derived anti-human TIGIT antibodies CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547, CHA.9.560, BM26, and BM29 increase IL-2 signaling. BM26 and BM29 belong to the mIgG1 isotype. The non-blocking human TIGIT antibody, CHA.9.543, does not enhance IL-2 signaling. Representative data (n>2) demonstrate the ESE (mean +/- standard deviation) of the luciferase signal from a 6-hour co-cultivation of human Jurkat IL-2-RE luciferase cells and human aARC CHOK1 PVR cells. The concentration of each antibody was 10 μg/ml.

Fig. 63. The results of Example 6 are shown showing that hybridoma-derived and reference anti-human TIGIT mIgG1 antibodies, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547, CHA.9.560, and BM26 increase IL-2 signaling in a dose-dependent manner. BM26 belongs to the mIgG1 isotype. Representative data (n>2) demonstrate the ESE (mean +/- standard deviation) of the luciferase signal from a 6-hour co-culture of human Jurkat IL-2-RE luciferase cells and human Cho-Κι PVR aARC cells. For each antibody, 10-point, 2-fold dilution series were used, starting at 20 μg/ml.

Fig. 64. Phage, hybridoma and BM antibodies against human TIGIT, CPA.9.027, CPA.9.049, CPA.9.059, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547, CHA.9.560, BM26 and BM29 are depicted increase antigen-specific IFN-γ signaling. BM26 is tested as both hIgG4 and mIgG1 isotypes, while BM29 is tested as hIgG4 isotype only. Representative data (n = 2) show the amount of IFN-γ (mean +/- standard deviation) in culture supernatant after 24 hours of co-cultivation of CMV-specific CD8+ T cells with human Mel624 PVR cells. The concentration of each antibody was 10 μg/ml. Mel624 human PVR used in the assay was injected with 0.0033 μg/ml or 0.001 μg/ml of the peptide.

Fig. 65. It is shown that phage, hybridoma and BM antibodies against human TIGIT, CPA.9.027, CPA.9.049, CPA.9.059, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547 and CHA.9.560, such as and BM26 enhance antigen-specific IFN-γ signaling either alone (blank bars) or in combination with the anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) (shaded bars). BM26 belongs to the mIgG1 isotype. For isotype control antibody therapies, the open bar refers to the isotype antibody only and the shaded bar refers to the isotype antibody in combination with CHA.7.518.1.H4 (S241P). Representative data (n = 2) show the amount of IFN-γ (mean +/- SD) in the culture supernatant after 24-hour co-culture of CMV-specific CD8+ T cells with Mel624 cells overexpressing human PVR and human PVRL2. The concentration of each antibody was 10 μg/ml. Human PVR/human PVRL2 Mel624 cells used in the assay were injected with 0.0033 or 0.001 μg/ml of peptide.

Fig. 66. Percentage increase in IFN-γ secretion using anti-human TIGIT antibody, CHA.7.518.1.H4 (S241P) and combination of anti-human TIGIT antibody and CHA.7.518.1.H4 (S241P) compared to the corresponding isotype control antibodies is shown. .

In FIG. 67 shows the dendrogram for the epitope-specific sorting experiments from Example 7.

In FIG. 68 shows a group of antibodies from the epitope-specific sorting experiments from Example 7.

Fig. 69. Depicted high affinity binding to cells overexpressing human TIGIT by dose titration of affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), controls. antibodies (BM26, BM29) and isotype control hIgG4 (anti-synagis) on Expi293 TIGIT cells overexpressing human TIGIT as described in the experiments of Example 3. /ml (133.33 nM [binding site]). AF647-labeled goat anti-F(ab') human (Jackson Immunoresearch) was added to cells to determine the binding of anti-TIGIT antibodies. Shown are gSIF of anti-TIGIT antibodies associated with Expi293 cells overexpressing human TIGIT (black line) and parental Expi293 cells (gray line). KD values +/- 95% CI and fit curves are shown below each graph.

Fig. 70. Anti-TIGIT antibodies are shown to cross-react with cynomolgus TIGIT when titrated with affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), reference antibodies (BM26, BM29) and hIgG4 isotype control (anti-synagis) on Expi293 cells overexpressing cynomolgus macaque TIGIT as described in the experiments from Example 3. ml (133.33 nM [binding site]). AF647-labeled goat anti-F(ab') human (Jackson Immunoresearch) was added to cells to determine the binding of anti-TIGIT antibodies. Shown are gSIF of anti-TIGIT antibodies associated with Expi293 cells overexpressing cynomolgus monkey TIGIT (black line) and parental Expi293 cells (grey line). KD values +/- 95% CI and fit curves are shown below each graph.

Fig. 71. Affinity matured phage antibodies are shown to cross-react with

- 11 040773 mouse TIGIT in dose titration of affinity matured phage antibodies reformatted into mouse IgG1 (mIgG1) (CPA.9.083, CPA.9.086), control anti-mouse TIGIT antibodies (BM27 mIgG1, BM30 mIgG1), and mIgG1 isotype control ( anti-synagis) as described in the experiments of Example 3. A) gSIF of anti-TIGIT antibodies associated with mouse TIGIT overexpressing HEK cells (black line) and parental HEK cells (grey line). C) hSIF anti-TIGIT antibodies (black line) or mIgG1 synagis (gray line) associated with regulatory CD4+CD25+Foxp3+ T cells isolated from s.c. implanted Renca tumors in Balb/c mice. Anti-TIGIT antibodies were titrated using a 2-fold or 3-fold dilution series starting at 15 μg/ml (200 nM [binding site]) or 10 μg/ml (132 nM [binding site]), respectively. AF'647-labeled goat anti-mouse IgG-Fc (Southern Biotech) was added to cells to detect anti-TIGIT antibody binding on cells overexpressing mouse TIGIT. Anti-TIGIT antibodies were directly conjugated to AF647 to bind mouse Tregs. The KD values for each anti-TIGIT antibody are shown.

Fig. 72. Dose titration of affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13) and control antibodies (BM26, BM29) for CD95+CD28-CD8 is shown. +CD3+ human memory effector T cells from PBMCs from 3 healthy donors (donors 321, 322 and 334) as described in the experiments of Example 3. PBMCs were stained on the surface with antibodies against the following lineage markers CD3, CD4, CD8, CD14, CD16, CD28, CD56 and CD95 (BD Biosciences, BioLegend), and live/dead fixable aqua dye (Life Technologies). AF647-labeled anti-TIGIT antibodies and an isotype control antibody hIgG4 (anti-synagis) were then titrated using a serial 3-fold dilution at 12 points starting at 30 μg/ml (396 nM [binding site]). Depicted are gSIF anti-TIGIT antibodies associated with effector memory T cells. The KD values for each antibody in 3 different donors are shown in the table. The affinity matured phage antibodies (CPA.9.083 and CPA.9.086) had the highest binding affinity to human effector memory T cells.

Fig. 73. Dose titration of affinity matured phage antibodies (CPA.9.083, CPA.9.086, CPA.9.103), humanized hybridoma antibody (CHA.9.547.1) and control antibody (BM26) on effector CD95+CD28-CD8+CD3+ T is shown. memory cells from PBMCs isolated from 2 naïve cynomolgus monkeys (BioreclamationIVT) as described in the experiments of Example 3. PBMCs were stained on the surface with antibodies against the following markers of the CD3, CD4, CD8, CD14, CD16, CD28, CD56 and CD95 lineages (BD Biosciences , BioLegend), as well as live/dead fixable aqua dye (Life Technologies). AF647 labeled anti-TIGIT antibodies and isotype control antibody hIgG4 (anti-synagis) were then titrated using a serial 3-fold dilution at 12 points starting at 30 μg/ml (396 nM [binding site]). GSIF of anti-TIGIT antibodies associated with effector memory T cells are shown with the subtraction of gsIF of the hIgG4 anti-synagis isotype control antibody. The KD values for each antibody from 2 different donors are shown in the table. The affinity matured phage antibodies (CPA.9.083 and CPA.9.086) had the highest binding affinity to cynomolgus monkey effector memory T cells.

Fig. 74. Depicted PIR binding of anti-TIGIT antibody to human, cynomolgus and mouse TIGIT as described in the experiments of Example 5. Kinetic rate and equilibrium dissociation constants for affinity matured phage antibodies (CPA.9.083, CPA.9.086, CPA.9.103) , humanized hybridoma antibodies (CHA.9.547.1 and CHA.9.547.7) and control antibodies (BM26, BM29) were determined using NIR on the ProteOn instrument.

Fig. 75. Anti-TIGIT antibodies are shown to block PVR/TIGIT interactions as described in the experiments of Example 4. Expi293 cells overexpressing human TIGIT were preincubated with either affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), control antibodies (BM26, BM29), or hIgG4 isotype control (anti-synagis). All antibodies were titrated using a serial 2.5-fold dilution at 11 points starting at 10 μg/ml (133.33 nM [binding site]). After pre-incubation of the antibody, human PVR-m2aFc was added to the cells at a concentration of 158 nM [binding site] or EC ₉₀ . AF647-labeled goat anti-mouse IgG-Fc (Southern Biotech) was then added to the cells to detect anti-TIGIT antibody binding. Percent inhibition of PVR-m2aFc binding to Expi293 cells overexpressing human TIGIT is shown for each antibody. IC ₅₀ values for each anti-human TIGIT antibody are shown in the table (n = 2 experiments).

Fig. 76. The results of Example 6 are shown, where affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13) and control antibody (BM26) increase IL-2 signaling. dose-dependent manner. Synagis hIgG4 is an isotype control antibody. Representative data (n>2) demonstrate the ESE (mean +/- standard deviation) of the luciferase signal from a 6-hour co-culture of Jurkat IL-2-RE human TIGIT luciferase cells and human Cho-Κι PVR cells. Series with 19 points, 1.5-fold dilution, starting at 20 μg/ml, were used for each antibody.

- 12 040773

Fig. 77. Anti-TIGIT antibodies have been shown to induce IFN-γ in CMV-specific CD8+ T cells. An in vitro co-culturing assay with human CMV-specific CD8+ T cells was used to assess the effect of affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13) and control antibodies (BM26, BM29) for antigen-specific cytokine secretion as described in the experiments of Example 6. The target cell line used in the assay was HLA-A2+ pancreatic adenocarcinoma cells, Raps.05.04, which endogenously express human PVR and PVRL2. CMV pp65 peptide was introduced into Raps.05.04 cells at a concentration of 0.03 μg/ml or 0.01 μg/ml at 37°C for 1 hour. The cells were then washed and plated at 50,000 cells/well in 96-well round bottom tissue culture plates. Human anti-TIGIT antibodies or hIgG4 isotype control antibody (anti-synagis) were added at a concentration of 0.1 μg/ml. Human CMV-specific CD8+ T cells from a single donor were expanded according to the above protocol. 50,000 human CD8+ T cells were added to each well. Co-cultures were incubated at 37° C. with 5% CO ₂ for 24 hours. The amount of human interferon gamma (IFN-γ) in the co-culture supernatant was measured by flow cytometry using a cytometric granule assay (BD Biosciences). The percentage increase in IFN-γ secretion for each antibody compared to the hIgG4 isotype is shown in the table (n=2 experiments).

Fig. 78. Anti-TIGIT antibodies are shown to increase IFN-γ in combination with an anti-PVRIG antibody, CHA.7.518.1.H4 (S241P). An in vitro co-culturing assay with human CMV-specific CD8+ T cells was used to assess the effect of affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13) and control antibodies (BM26, BM29) for antigen-specific cytokine secretion in combination with an anti-PVRIG antibody, CHA.7.518.1. The target cell line used in the assay was HLA-A2+ pancreatic adenocarcinoma cells, Raps.05.04, which endogenously express human PVR and PVRL2. CMV pp65 peptide was introduced into Raps.05.04 cells at a concentration of 0.03 μg/ml or 0.01 μg/ml at 37°C for 1 hour. Then the cells were washed and placed in an amount of 50,000 cells/well in 96-well round bottom plates for tissue culture. Human anti-TIGIT antibody or hIgG4 isotype control antibody (anti-synagis) was added at a concentration of 0.1 μg/ml in combination with CHA.7.518.1 (shaded bars) or hIgG4 isotype control antibody at 10 μg/ml (solid bars) . Human CMV-specific CD8+ T cells from a single donor were expanded according to the above protocol. 50,000 human CD8+ T cells were added to each well. The co-cultures were incubated at 37° C. with 5% CO ₂ for 24 hours. The amount of human IFN-γ in the co-culture supernatant was measured by flow cytometry using a cytometric granuloassay (BD Biosciences). The percentage increase in IFN-γ secretion for each antibody compared to the hIgG4 isotype is shown in the table (n=2 experiments).

Fig. 79. Correlation analysis of PVRIG and TIGIT expression on CD4+ and CD8+ T cells from dissociated tumors is shown. For each tumor sample, the mean fluorescence intensity ratio (MFIr) was calculated and Spearman's correlation analysis was performed, and the r2 and p values are indicated.

Fig. 80. Results of tumor growth inhibition and survival in TIGIT knockout mice treated with an anti-mouse PVRIG antibody are shown. Groups of 7-10 TIGIT and C57BL/6 DT knockout mice were injected s.c. with 1 x 105 B16/Db-hmgp100 cells. Mice were treated with the designated antibody twice a week for 3 weeks starting on the day of inoculation (day 0). A) Mean tumor volumes +/- standard error of the mean (SEM) are shown in the upper plot, where *** denotes the p value <0.001 for TIGIT knockouts treated with anti-murine PVRIG antibody (clone 407) compared to C57BL/6 DT treated with control antibody mIgG1. Tumor volumes for individual mice in each antibody treatment group are shown as radar charts in the lower graphs. B) Table that describes the TGI measured on the indicated days compared to C57BL/6 DT control mice treated with the mIgG1 isotype. C) Survival of mice after s.c. injection of B16/Db-hmgp 100 cells.

In FIG. 81A-C depict combination therapies with these antibodies compared to controls in Mel-624, Colo205 and Raps.05.04 cells. gp100 or CMVpp65-specific T cells were co-cultured with Mel-624, Colo205 and Raps.05.04 cells, gp100 or CMVpp65 peptide and the indicated antibodies at a concentration of 10 mg/ml. The concentration of IFN-γ in conditioned media was determined after 24 hours. Depicted is the mean + standard. off of three values. The % change in IFN-γ for each condition relative to hIgG4 is shown.

In FIG. 82A-C depict PD-1/TIGIT/PVRIG expression on CD8 T cells and PD-L1, PVR, PVRL2 expression on Colo205, Panc.05.04 cells. A) Expression of PVRIG, TIGIT and PD-1 on CMVpp65 reactive T cells expanded with pp65 peptide with IL-2 and IL-7 for 10 days. The expression of PVRIG, TIGIT and PD-1 on CMVp65-reactive T cells is shown. B) Depicted Express

- 13 040773 PD-L1, PVR and PVRL2 on Co10205 and Raps cells. 05.04. C) CMVpp65-specific T cells were co-cultured with Colo205 and Raps.05.04 cells, CMVpp65 peptide and the indicated antibodies at a concentration of 10 mg/ml. The concentration of IFN-γ in conditioned media was determined after 24 hours. Displayed mean + std. off of three values. The % change in IFN-γ for each condition relative to hIgG4 is shown.

Fig. 83. PVRIG is expressed most highly in subpopulations of cytotoxic lymphocytes from human cancer. A) PVRIG expression is depicted in leukocyte subpopulations from PBMCs from 5-8 healthy donors. PVRIG expression is defined as the ratio of PVRIG SIF to isotype control SIF. B) Expression of PVRIG, TIGIT, CD96, and PD-1 on peripheral blood Treg compared to CD8 T cell subpopulations from PBMCs from 5 healthy donors. C) CMV pp65-specific T cells from 3 healthy donors were expanded in vitro with pp65(495-503) peptide, IL-2 and IL-7 over a period of up to 7 days. Shown is the expression of TIGIT (blue) and PVRIG (black) on HLA-A2/pp65 (495-503) on tetramer positive cells. D) Human T cells were cultured with allogeneic DCs and on days 0, 1, 2 and 7 after activation TIGIT and PVRIG expression is shown on CD4+ T cells. E) Representative FACS plots showing PVRIG expression (blue) versus isotype control (red) for TILs (CD4 T cells, CD8 T cells and NK cells) from representative lung and kidney cancers. F) Co-expression of PVRIG, TIGIT and PD-1 on CD4 and CD8 TIL from a lung cancer sample is shown. G) PVRIG expression on CD8+ and CD4+ TILs from dissociated human tumors of various cancer types is depicted. Each dot represents a different tumor in a different patient. H) Relative expression on CD8 TIL against Treg TIL for PVRIG, TIGIT and PD-1 from endometrial, kidney and lung tumors was assessed. For each tumor, CD8 TIL fold expression was normalized to Treg TIL fold expression and plotted. For A, B, C, G, and H, the mean ± SOS is shown with error bars.

Fig. 84. PVRL2 expression is enhanced in the tumor microenvironment. A) PVRL2 expression was assessed by IHC in tumors of the lung, ovary/endometrium, breast, colon, kidney, and melanoma. Bars represent mean ± SOS. For each tumor, 2 biopsies were evaluated by a pathologist and quantified by the extent and intensity of membrane staining on tumor cells as described in Supplementary Methods. For each tumor, the average score of 2 biopsies is shown. B) Depicted is a representative melanoma tumor showing PVRL2 expression on tumor cells (arrow) and on immune cells (*) in the stroma. C) Depicted PVRL2 expression on the Iog2 scale from dissociated tumors determined by FACS on subpopulations of CD45 ^- , CD14+ TAM and Lin ^- CD14 ^- CD33 ^hl mDC (myeloid dendritic cells) cells. The mean ± SOS for each type of cancer is shown. The dotted line means no staining. For each cell type, at least 100 events were needed for analysis. D) Representative FACS plots of PVRL2 expression (blue) versus IgG (red) are shown for lung cancer. E) For tumor samples where we were able to evaluate both PVRIG and PVRL2 expression, PVRIG expression on CD8+ T cells is plotted against PVRL2 expression on CD14+ and CD45 ^TAMS cells for each tumor. Each dot represents a separate tumor sample. The red line represents 2-fold expression of PVRIG or PVRL2 compared to IgG. On the table in Fig. 84F demonstrates the predominance of PVRL2 in various tumor samples.

Fig. 85. Different regulation of PVRL2 and PD-L1 on tumor cells. A) Expression of PD-L1 and PVRL2 was assessed by IHC in serial sections. Tumor samples from Fig. 84A were grouped by tissue type and depict PVRL2 expression on PD-L1 negative and PD-L1 positive. PDL1-negative tumors were defined as those that did not have tumor membrane staining or immune cells in both two identical biopsies for a given tumor. PD-L1 positive staining was defined as membrane staining on at least 1 of the tumor biopsies. Bars represent the mean ± SOS for each group. B, C) Representative expression of PVRL2+ PD-L1 ^- endometrial tumors (B) and PVRL2+ PD-L1' lung tumors (C). D) Immature BM-DCs (dendritic cell of bone marrow origin) were cultured with the indicated stimuli and the expression of PVR, PVRL2 and PD-L1 was assessed by FACS on the 2nd day of culture. For each condition, expression was normalized to medium-only control conditions. E) Expression of PVR, PVRL2 and PD-L1 is shown on HT-29 cells treated with IFN-γ or media alone. PD-L1 or PVRL2 are depicted in blue, and staining using IgG isotype control is depicted in red.

Fig. 86. CHA.7.518.1.H4 (S241P) is a high affinity antibody that enhances T cell activation. A) Shows binding of CHA.7.518.1.H4 (S241P) or IgG isotype control to HEK293 PVRIG or HEK293 parent cells by FACS. For CHA.7.518.1.H4 (S241P) binding to HEK293 hPVRIG, HEK293 cPVRIG and Jurkat cells, KD values determined by FACS are shown. C) CHA.7.518.1.H4 (S241P) disrupts PVRL2 Fc binding to HEK293 cells ectopically expressing PVRIG. Average value shown ± std. off of three values. C) CHA.7.518.1.H4 (S241P) blocks the binding of PVRIG Fc to HEK293 cells, which en

- 14 040773 pregenously express PVRL2. D) Human CD4 T cells were co-cultured with CHO iAPK cells expressing anti-CD3 antibodies that bind to the cell surface and hPVRL2 in the presence of 10 μg/ml of anti-PVRIG antibodies and isotypic human IgG antibodies. The effect of anti-PVRIG antibodies on the proliferation of CD4 T cells isolated from 11 different donors is shown. Bars represent mean ± SOS. E) gp100-specific T cell lines (TIL-209, TIL-463) were co-cultured with CHO cells modified to express HLA-A2 and PVRL2, together with 10 μg/ml anti-PVRIG or IgG isotype control antibody. The production of IFN-γ and TNF-α was checked 24 h after co-cultivation. Average value shown ± std. off from three values. The percentage change in IFN-γ and TNF-α for each condition relative to the isotype control is shown by a number above each bar. F) Expression of PVR, PVRL2 and PD-L1 (red) relative to IgG (blue) on MEL624, Colo205 and Panc.05.04 cells. T cells show expression of PVRIG, TIGIT and PD-1 (red) relative to IgG (blue) on TIL-209 and TIL-463 gp100-specific T cells as well as on CMVpp65-specific T cells. To propagate CMVpp65-reactive T cells, PBMCs were cultured with pp65 (495-503) peptide, IL-2 and IL-7 for 10 days. Expression of PVRIG, TIGIT, PD-1 shown on HLA-A2/pp65 tetramer-positive cells G) gp100-specific T cells (TIL-209, TIL-463) propagated from TIL derived from melanoma tumors were co-cultured with cells MEL624 in the presence of 10 μg/ml of the indicated antibodies. The concentration of IFN-γ in the conditioned media was determined after 24 h. H, I) Expanded CMVpp65-specific T cells were co-cultured with Colo205 and Raps.05.04 cells, CMVpp65 peptide and the indicated antibodies at 10 μg/ml. The concentration of IFN-γ in conditioned media was determined after 24 hours. For E, G, H, I, the mean value ± standard is shown. off from three values. The percentage change in IFN-γ for each condition relative to the isotype control is shown by a number above each bar.

Fig. 87. PVRIG-deficient mice have increased T-cell function. A) PVRIG RNA expression was measured by qRT-PCR (Quantitative Real Time PCR) from purified mouse immune cell subsets. Relative expression for housekeeping genes was determined by the ACt method. IN). Transgenic pmel CD8 ⁺ TCR T cells were activated with gp100 (25-33) and PVRIG and TIGIT RNA transcript levels assessed by qRT-PCR at the indicated time points. The graph shows the mean ± SOS of the results from 5 different experiments. C) Spleens were harvested from PVRIG ^- / ^- and DT littermates and analyzed by flow cytometry for PVRIG expression on NK, CD4+ and CD8+ T cells (resting cells). In addition, CD3+ T cells were isolated from splenocytes and activated for 11 days with anti-CD3/anti-CD28 beads. After activation, PVRIG expression on CD4+ and CD8+ T cells (activated cells) was analyzed by flow cytometry. Each dot represents cells from an individual mouse. D) Splenocytes with DT and PVRIG ^- / ^- were labeled with eFluor450 cell proliferation dye and cultured in the presence of control-Fc (mouse IgG2a) or mouse PVRL2 Fc. After 4 days of cultivation, cell division was analyzed by flow cytometry. Shown are the results of the FACS analysis from the experiment (left) and the total percentage of PVRL2 Fc inhibition (defined as % control-Fc proliferation subtracted from % PVRL2 Fc proliferation). 3 independent experiments are shown (right). * indicates p-value <0.05, paired Student's t test for change in proliferation in the presence of PVRL2-FC relative to proliferation in the presence of a control protein in DT compared to PVRIG ^-/- T cells. E) pmel CD8+ T cells derived from pmel PVRIG ⁻ / ⁻ or pmel PVRIG DT mice were activated for 11 days with their cognate peptide and IL-2. The pmel-activated CD8+ cells were then cultured with B16Db/gp100 cells for 18 hours and, after co-culture, evaluated for CD107 expression and cytokine production. Four independent experiments are presented, as indicated by each paired point. * indicates p-value <0.05, Student's t-test comparing PVRIG ^- / ^- and DT.

Fig. 88. Deficiency of PVRIG leads to a decrease in tumor growth and an increase in the effector mechanism of CD8+ T cells. A) C57BL/6 DT or PVRIG ^- / ^- mice were injected subcutaneously with 5 x 105 MC38 cells. Tumor volumes were measured twice a week, n = 10 mice per group, mean ± SOS is shown, * indicates p-value <0.05 according to unpaired Student's t-test for TD mice vs. PVRIG ^- / ^- mice (ANOVA). C) Individual tumor growth curves are shown, n = 10 mice per group, one representative experiment is shown (n = 2). C) C57BL/6 DT or PVRIG ⁻ / ⁻ mice were injected subcutaneously with 5α105 MC38 cells. On the 14th day after inoculation, mice were treated with anti-PD-L1, 2 times a week for 2 weeks. Tumor volumes were measured 2 times per week, n = 10 mice per group, showing mean ± SOS, p-value = 0.052 by unpaired Student's t-test for TD mice against PVRIG ^- / ^- mice treated with anti-PD-L1. D) Individual tumor growth curves are shown. One representative experiment is depicted (n = 2). E-H) In separate replicates, tumors were harvested on day 18 after mice received 2 doses of anti-PD-L1 or an appropriate isotype control. Dissociated tumors were enriched for CD45+ cells prior to 4 h stimulation with PMA and ionimicin in the presence of Brefeldin A. In the graphs

- 15 040773 shows the total numbers per mg of tumor tissue of CD45+ immune cells, CD8+ T cells and interferon-γ producing CD8+ T cells from isotype-treated wild-type and PVRIG'' mice (E) and from wild-type and PVRIG' mice /' treated with anti-PD-L1 (F). GH). The frequency of CD8+ IFN-a ⁺ ΦΗΟ-γ ⁺ effector cells in tumor-draining lymph nodes from PVRIG'/' mice treated with isotype and anti-PD-Ll is shown, compared to their respective wild-type cohort. For E-H, means ± SOS are shown and p-values from unpaired Student's t-test are shown.

Fig. 89. Antagonistic anti-PVRIG antibodies synergistically inhibit tumor growth in combination with PD-1 inhibitors or TIGIT genetic deficiency. A) The mPVRL2 Fc fusion protein was shown to bind to modified mVVRIG HEK293 cells that were pre-incubated with serial dilutions of anti-mPVRIG mAb or IgG isotype control Ab. B) BALB/c mice were injected subcutaneously with 5 x 105 CT26 cells. On the 14th day after inoculation, mice were sacrificed, and the spleen, draining lymph nodes and tumors were collected. Cells were analyzed by flow cytometry for PVRIG expression on CD3 ⁺ CD4+ T cells, CD3 ⁺ CD8+ T cells, CD3'CD49b ⁺ NK cells, CD11b ⁺ Gr-1 ⁺ myeloid suppressor cells (MLSC) and CD11b ⁺ F4/80 + macrophages. C, D) BALB/c mice were injected subcutaneously with 5x105 CT26 cells. On the 7th day after inoculation, mice were treated with anti-PD-Ll and/or anti-PVRIG Ab, 2 times a week for 3 weeks (arrows indicate treatment with Ab). C) Tumor volumes are shown. *** indicates p-value <0.001 (ANOVA) for rat aPD-Ll + IgG2b compared to α groups treated with PD-L1 + aPVRIG. Arrows indicate antibody dosing. D. Complete response mouse survival analysis. * indicates p-value <0.05 (log rank test) for α PD-L1 + IgG2b rat compared to aPD-L1 + aPVRIG treated groups. One representative study of 3 studies is depicted. E. C57BL/6 or TIGIT'/' mice were injected subcutaneously with 1 x 105 B16/Db-hmgp100 cells. Mice were treated twice a week for 3 weeks with the assigned mAb starting on the day of inoculation (day 0). E. Tumor volumes were measured 2 times a week, and the mean ± SOS is shown. Tumor growth inhibition measured on indicated days compared to WT control + mIgG1 isotype control. *** indicates p-value <0.001 for TIGIT'/'+ aPVRIG compared to DT+ mIgG1 isotype control. Arrows indicate antibody dosing. F. Individual tumor growth curves for each mouse are shown. One representative experiment out of 2 performed is depicted.

Fig. 90. PVRIG is expressed on TIL and NK TIL cells in human cancer. A) Expression of PVRIG, TIGIT, CD96 and PD-1 on CD4 T cells from healthy donor PBMCs is shown. Shown is the mean value ± SOS. B) Human T cells were co-cultured with allogeneic PBMC and PVRIG protein expression on CD4 and CD8 T cells (top). C) Tumors were separated and individual cells were activated using anti-CD3 and anti-CD28. PVRIG expression (blue) relative to IgG isotype control (red) was assessed on day 0 (directly ex vivo) and on day 5 after activation. D) PVRIG expression on NK cells from dissociated human tumors is shown. Each dot represents a different tumor in a different patient. Shown is the mean ± 95% confidence interval. D) Dissociated tumor cells were activated using anti-CD3 and anti-CD28 beads for 5 days. Expression of PVRIG (blue) relative to IgG control (red) on CD4 and CD8 T cells on day 0 directly ex vivo and on day 5 post-activation is shown for 2 dissociated tumor samples. E) PVRIG expression was assessed on CD4 and CD8 T cells from dissociated tumors and from dissociated corresponding donor normal adjacent tissue. Each line represents the respective tissues obtained from an individual patient. Paired Student's t-test was performed. F) Shown is correlation analysis of PVRIG, TIGIT and PD-1 fold expression versus IgG isotype control on CD4 and CD8 T cells from tumors. Each dot represents a separate tumor sample. Spearman's correlation coefficient and p-value are shown.

Fig. 91. PVRL2 expression is increased in oncological diseases of the colon, skin and breast. A) Photomicrographs showing binding of anti-human PVRL2 antibody Sigma to FFPE (formalin-fixed and paraffin-embedded) sections of positive cells, human CHO-S PVRL2 (right) compared to negative cells, CHO-S (left), after antigen demasking at pH9. B) Anti-PVRL2 antibody was tested on a panel of PVRL2+ (HT29, MCF7, PC3, PANC1, RT4, NCI-H1573) and PVRL2' (Jurkat, OPM2, Daudi, CA46) cell lines. C-F) An example of PVRL2 expression in normal and cancerous lung tissues. C) Normal tissue that does not show staining. D) Adenocarcinoma of the lung, which show partial positive staining. E) Adenocarcinoma of the lung, which show positive staining. F) Adenocarcinoma of the lung, which show strong positive staining.

Fig. 92. PVRL2 is elevated on TAM and CD45' cells in tumor compared to normal adjacent tissue. Depicted is PVRL2 expression on CD45' cells and TAMs from a matched donor tumor and normal adjacent tissue. The p-value of the paired Student's t-test is shown.

Fig. 93. PVRIG and PVRL2 are co-expressed in the same tumor sample. Express

- 16 040773 PVRIG expression on CD4 T cells (A) and NK cells (B) is plotted against PVRL2 expression on TAMs for a single tumor.

Fig. 94. Activity of CHA.7.518.1.H4 (S241P) on human T cells. A) Expression of PVRIG on CD4 T cells activated by CHO cells expressing anti-CD3 and PVRL2 that bind to the cell surface. C) Expression of HLA-A2, B-2m and PVRL2 is shown on the original CHO-S and modified CHO-S cell lines. The fold expression relative to the isotype is represented by a number. C) CHO cells ectopically expressing cell surface bound anti-CD3 and PVRL2 were co-cultured with purified CD8 T cells in the presence of various concentrations of anti-PVRIG Ab or appropriate IgG control. The percentage of proliferation is shown. Each point represents the average of the three values. D) CHO cells that ectopically express HLA-A2/B2m and PVRL2 were co-cultured with 2 gp100-specific T cell lines (TIL F4, TIL 209) in the presence of 1 μg/ml gp100 and various concentrations of anti-PVRIG antibody or the corresponding IgG control. The concentrations of TNF-α on the 3rd day of co-cultivation are reduced. Each point represents the average of the three values.

Fig. 95. Characterization of binding interactions between mPVRIG and a surrogate anti-mPVRIG antibody. A, B) The binding of mPVRIG to mPVRL2 was assessed using surface plasmon resonance. C) Soluble Fc receptor or control proteins were incubated in a dose dependent manner with immobilized mPVRL2 HIS in ELISA format. The associated Fc receptor is depicted. D) Soluble PVRL2 HIS protein was dose-dependently incubated on PVRIG Fc or DNAM Fc coated plates. E) Binding of mPVRIG Fc or control Fc fusion protein to B16F10 cell line transfected with mPVRL2 siRNA, mPVR siRNA or transfected with scrambled siRNA is shown. F) Affinity characterization of the rat mAb against mouse PVRIG was performed by studying the binding of anti-mPVRIG to HEK293 cells overexpressing mPVRIG. G) Affinity characterization of a rat mAb against mouse PVRIG is shown by examining the binding of anti-mPVRIG to a D10.G4.1 cell line endogenously expressing mPVRIG against an isotype control of rat IgG. H) Anti-mPVRIG binding to D10.G4.1 cells transfected with mouse PVRIG siRNA (green bar graph) against scr siRNA (orange bar graph). I) Binding of mPVRIG Fc preincubated with anti-mPVRIG Ab to B16-F10 cells that endogenously express PVRL2

Fig. 96. Creation of PVRIG and TIGIT knockout transgenic mice. Conditional PVRIG knockout and Tigit knockout mouse strains were established at Ozgene Pty Ltd (Bentley, Western Australia, Australia). A) A targeting construct in which PVRIG exons 1 to 4 were flanked was electroporated into the C57BL/6 ES cell line, Bruce 4 (Koentgen et al, Int Immunol 5: 957-964, 1993). C) A targeting construct in which the Tigit 1 exon coding region (including ATG) and exons 2 and 3 were replaced with a FRT-flanked neo cassette was electroporated into the C57BL/6 ES, Bruce4 cell line. Homologous recombinant ES cell clones were identified by Southern blot hybridization and injected into goGermline blastocysts (Koentgen et al., genesis 54: 326-333, 2016). Male chimeric mice were generated and bred to C57BL/6J females to create heterozygous germline progeny with a C57BL/6 background. Germline mice were bred to the common FLP C57BL/6 mouse line to remove the FRT-flanked selectable marker cassette and generate conditional or knockout alleles (for PVRIG and Tigit, respectively). For the PVRIG knockout, the mice were then crossed with the common Cre C57BL/6 mouse line to remove 1oxP flanked exons and the knockout allele was generated.

Fig. 97. PVRIG knockout mice are immunophenotypically similar to wild-type mice. Mice (n=5 for wild-type and PVRIG cohorts) were euthanized before venous blood was drawn into anticoagulant-coated tubes and organs were harvested. Individual cells were isolated from freshly collected bone marrow, thymus, spleen, skin and mesenteric lymph nodes. Cells were stained with fluorochrome-conjugated surface marker antibodies and read on a BD LSR Fortessa flow cytometer. Panels show comparable frequencies of myeloid cells (A), dendritic cells (B), B cells (C), T cells (D), CD4 T cells (E), CD8 T cells (F), and NK cells. (G) by type of lymphoid tissue. (H-I) Venous whole blood was passed through the Hemavet 950 Veterinary Hematology System to compare differential scores and blood cell subtype frequencies in WT and PVRIG-deficient mice.

Fig. 98. Increased T cell effector function in anti-PDL1 treated PVRIG ⁻ / ⁻ mice compared to anti-PD-L1 treated DT mice. MC38 tumors were inoculated into DT or PVRIG ⁻ / ⁻ mice and subsequently treated with anti-PD-L1 or rat IgG2b isotype control. On day 18, CD45+ tumor-infiltrating lymphocytes were isolated from tumors, RNA was extracted, and transcript profiling was performed. Several genes associated with T cells are depicted, with each dot representing an individual mouse. Shown are p-values for Student's t-test.

Fig. 99. Anti-TIGIT and anti-PVRIG antibodies induce tumor cell death. In vitro co-culture assay with expanded human CMV-specific CD8+ T

- 17 040773 cells was used to evaluate the effect of control anti-TIGIT antibody and CHA.7.518.1.H4 (S241P) on antigen-specific killing of tumor cells. The HLA-A2+ target cell lines used in the assay were Mel624 (A) and Raps05.04 (B). Synagis hIgG4 is an isotype control antibody. Luciferase activity in target cells was measured using Bio-Glo luciferase substrate. Representative data (n>2) show percent specific kill (mean +/- SD) of Mel624 or Panc05.04 cells after 16 hours of co-culture with human CMV-specific CD8+ T cells from three different donors.

Fig. 100. Dose-dependent killing of tumor cells by anti-TIGIT antibodies using CHA.7.518.1.H4 (S241P). An in vitro co-culture assay with human CMV-specific CD8+ T cells was used to evaluate the effect of two different anti-TIGIT antibodies, BM26 and CPA.9.086, in combination with CHA.7.518.1.H4 (S241P) on antigen-specific killing of Mel624 cells. Luciferase activity in target cells was measured with Bio-Glo luciferase substrate. Representative data (n>2) show percent specific kill (mean+/sd) of Mel624 cells after 16 hours co-culture with human CMV-specific CD8+ T cells from a single donor.

Fig. 101. CDR sequences CPA.9.086, IMGT and Kabat numbering.

Fig. 102. The combination of anti-TIGIT hIgG4 + CHA.7.518.1.H4 (S241P) induces the destruction of tumor cells. Co-cultivation of CMV-reactive CD8+ T cells with Mel624 PVR, PVRL2 and a single dose of luciferase OE 10 μg/ml aTIGIT Ab I 10 μg/ml CHA.7.518.1.H4 (S241P) with CMV-reactive donor 4, while the dose was titrated starting with 0.5 µg/ml aTIGIT Ab and 10 µg/ml CHA.7.518.1.H4 (S241P) with CMV-reactive donor 156.

V. Detailed Description of the Invention

A. Overview

The present invention provides a number of useful antibodies for use alone or in combination in the treatment of cancer. Cancer can be thought of as the inability of the patient's body to recognize and eliminate cancer cells. In many cases, these transformed (eg, cancerous) cells oppose immune surveillance. There are natural control mechanisms that limit T cell activation in the body to prevent unrestricted T cell activity that can be used by cancer cells to evade or suppress the immune response. Restoring the ability of immune effector cells, especially T cells, to recognize and destroy cancer is the goal of immunotherapy. The field of immuno-oncology, sometimes referred to as immunotherapy, is rapidly evolving, with several recently approved antibodies that inhibit T cell checkpoints such as Yervoy, Keytruda, and Opdivo. These antibodies are commonly referred to as checkpoint inhibitors because they normally block the negative regulators of T-cell immunity. As a rule, it is understood that various immunomodulatory signals, both co-stimulatory and co-inhibitory, can be used to control the optimal antigen-specific immune response. Typically, these antibodies bind to checkpoint inhibitor proteins such as CTLA-4 or PD-1, which under normal circumstances prevent or suppress the activation of cytotoxic T cells (CTLs). By inhibiting the checkpoint protein, for example through the use of antibodies that bind these proteins, an enhanced T cell response against tumors can be achieved. That is, these cancer checkpoint proteins suppress the immune response; when the proteins are blocked, for example, using antibodies to the checkpoint protein, the immune system is activated, resulting in immune stimulation, resulting in the treatment of pathological conditions such as cancer and infectious disease.

This invention is directed to the use of antibodies to additional checkpoint proteins, PVRIG and TIGIT. PVRIG is expressed on the cell surface of NK and T cells and shares several similarities with other known immune checkpoints. The identification and methods used to show that PVRIG is a checkpoint receptor are discussed in WO 2016/134333, expressly incorporated herein by reference. Anti-human PVRIG antibodies that block PVLR2 interaction and/or binding are provided herein. When PVRIG binds to its ligand (PVRL2), an inhibitory signal is generated that acts to attenuate the immune response of NK and T cells against the target cell (ie similar to PD1/PDL1). Blocking PVRL2 binding to PVRIG turns off this PVRIG inhibitory signal and, as a result, modulates the immune response of NK and T cells. The use of an anti-PVRIG antibody that blocks binding to PVRL2 is a therapeutic approach that enhances the killing of cancer cells by NK and T cells. Blocking antibodies have been created that bind PVRIG and block the binding of its ligand, PVRL2. Anti-PVRIG antibodies are presented in combination with other checkpoint inhibitor antibodies, such as PD-1.

Similarly, TIGIT has also been shown to have characteristic checkpoint receptor properties and the present invention provides anti-TIGIT antibodies that block the interaction

- 18 040773 and/or linking TIGIT to PVR. When TIGIT is bound by its ligand (PVR), an inhibitory signal is generated that acts to attenuate the immune response of NK and T cells against the target cell (ie similar to PD-1/PDL1). Blocking PVR binding to TIGIT turns off this TIGIT inhibitory signal and, as a result, modulates the immune response of NK and T cells. The use of an anti-TIGIT antibody that blocks PVR binding is a therapeutic approach that enhances NK and T cell killing of cancer cells. Blocking antibodies have been created that bind TIGIT and block the binding of its ligand, PVR. Anti-TIGIT antibodies are provided in combination with other checkpoint inhibitor antibodies, such as PD-1.

In addition, the invention provides combinations of anti-PVRIG antibodies and anti-TIGIT for use in the treatment of cancer.

B. Definitions

In order to be able to more fully understand this application, several definitions are set forth below. Such definitions are intended to cover grammatical equivalents.

Ablation in this document means the reduction or exclusion of activity. In some embodiments, it is useful to remove activity from antibody constant domains. Thus, for example, FcyR binding ablation means that the amino acid variant of the Fc region has less than 50% of the original binding compared to the Fc region that does not contain the specific variant, preferably with less than 70-80-90-95-98% loss of activity, and generally below the level of detectable binding in the Biacore assay. As shown in FIG. 50, one ablation variant in the IgG1 constant region is the N297A variant, which removes the native glycosylation site and significantly reduces FcyRIIIa binding and thus reduces antibody-dependent cell-mediated cytotoxicity (ADCC).

By antigen-binding domain or ABD is herein meant a set of six Complementary Determining Regions (CDRs) that, when present as part of a polypeptide sequence, specifically bind the target antigen as discussed herein. Thus, the TIGIT antigen-binding domain binds the TIGIT antigen (the sequence of which is depicted in Fig. 51) as described herein. Similarly, the binding domain of an anti-PVRIG antibody binds the PVRIG antigen (the sequence of which is depicted in Figure 1) as described herein. As known in the art, these CDRs are typically present as a first set of heavy chain variable region CDRs (vhCDR or VH CDR) and a second set of light chain variable region CDRs (VLCDR or _Vl CDR), each containing three CDRs: vhCDR1 , vhCDR2, vhCDR3 for the heavy chain and VlCDR1, VLCDR2 and VLCDR3 for the light chain. The CDRs are present in the heavy chain variable region and light chain variable region domains, respectively, and together form the Fv region. Thus, in some cases, the six CDRs of the antigen-binding domain are introduced via the heavy chain variable region and the light chain variable region. In Fab format, a set of 6 CDRs supplemented with two different polypeptide sequences, a heavy chain variable domain (vh or VH containing vhCDR1, vhCDR2 and vhCDR3) and a light chain variable domain (vl or Vl containing VlCDR1, VLCDR2 and VLCDR3), where The C-terminus of the vh domain is attached to the N-terminus of the heavy chain CH1 domain, and the C-terminus of the vl domain is attached to the N-terminus (and thus forms the light chain).

Modification herein means an amino acid substitution, an insertion and/or deletion in a polypeptide sequence, or a change in a moiety chemically linked to a protein. For example, the modification may be by changing the carbohydrate or PEG structure attached to the protein. Amino acid modification as used herein means an amino acid substitution, insertion and/or deletion in a polypeptide sequence. For clarity, unless otherwise indicated, amino acid modification always refers to an amino acid that is encoded by DNA, eg, the 20 amino acids that have codons in DNA and RNA.

Amino acid substitution or substitution herein means the replacement of an amino acid at a particular position in the primary polypeptide sequence with another amino acid. In particular, in some embodiments of the invention, the substitution is an amino acid that does not occur naturally in a particular position, does not occur naturally in this organism, or in any organism. For example, the N297A substitution refers to a variant polypeptide in which the Fc variant in which glutamic acid is at position 297 is replaced with a tyrosine. To be clear, a protein that is engineered to change the coding nucleotide sequence but without changing the original amino acid (e.g., changing CGG (coding for arginine) to CGA (also encoding for arginine) to increase expression levels in the host) is not a protein with an amino acid substitution ; that is, despite the creation of a new gene encoding the same protein, if the protein has the same amino acid at the particular position it starts at, then it is not an amino acid substitution.

By amino acid insertion or insertion, as used herein, is meant the addition of an amino acid sequence at a specific position in the original polypep.

- 19 040773 tid sequence. For example, -233E or 233E indicates the insertion of glutamic acid after position 233 and before position 234. In addition, -233ADE or A233ADE indicates the insertion of AlaAspGlu after position 233 and before position 234.

By amino acid deletion, or deletion as used herein, is meant the removal of an amino acid sequence at a specific position in the original polypeptide sequence. For example, E233- or E233#, E233() or E233del denotes a deletion of glutamic acid at position 233. In addition, EDA233- or EDA233# denotes a deletion of the GluAspAla sequence that starts at position 233.

By variant protein or protein variant or variant, as used herein, is meant a protein that differs from the protein of the parent protein by at least one amino acid modification. A protein variant may refer to the protein itself, the composition containing the protein, or the amino acid sequence that encodes it. Preferably, the protein variant has at least one amino acid modification compared to the original protein, for example, from about one to about seventy amino acid modifications, and preferably from about one to about five amino acid modifications compared to the original. As described below, in some embodiments, the parent polypeptide, e.g., the parent Fc polypeptide, is a wild-type human sequence, such as an Fc region from IgG1, IgG2, IgG3, or IgG4, although variant human sequences can also serve as parent polypeptides. The protein herein will preferably have at least about 80% identity with the sequence of the parent protein, and most preferably at least about 90% identity, more preferably at least about 95-98-99% identity. Variant protein can refer to the variant protein itself, compositions containing the variant protein, or the DNA sequence that encodes it. Accordingly, an antibody variant or variant antibody as used herein means an antibody that differs from the parent antibody due to at least one amino acid modification, an IgG variant or variant IgG as used herein means an antibody that differs from the parent IgG (again, in many cases, from the sequence human IgG) due to at least one amino acid modification, and a variant immunoglobulin or variant immunoglobulin as used herein means an immunoglobulin sequence that differs from such sequence of the parent immunoglobulin due to at least one amino acid modification. Fc variant or Fc variant as used herein means a protein containing an amino acid modification in the Fc domain. Fc variants according to the invention are defined according to the amino acid modifications that constitute them. For example, S241P or S228P is a hinge variant with a proline substituted at position 228 relative to the original hinge IgG4 polypeptide, with S228P numbering corresponding to the EC index and S241P corresponding to Kabat numbering. The EC index, or EC-index, as in the Kabat or EC numbering scheme, refers to the numbering of an EC antibody (Edelman et al, 1969, Proc Natl Acad Sci USA 63:78-85, incorporated herein by reference in its entirety) . The modification may be an addition, a deletion, or a substitution. Substitutions may include natural amino acids and, in some cases, synthetic amino acids. Examples include US Pat. Nos. 6,586,207; W098/48032; WO 03/073238; US2004-0214988A1; WO 05/35727A2; WO 05/74524A2; J. W. Chin et al., (2002), Journal of the American Chemical Society 124: 9026-9027; J. W. Chin, & P. G. Schultz, (2002), ChemBioChem 11: 1135-1137; J. W. Chin, et al., (2002), PICAS United States of America 99:11020-11024; and L. Wang, & P. G. Schultz, (2002), Chem. 1-10 are all incorporated by reference in their entirety.

As used herein, protein means at least two covalently attached amino acids, which include proteins, polypeptides, oligopeptides, and peptides. The peptidyl group may contain naturally occurring amino acids and peptide bonds or synthetic peptidomimetic structures, i.e. analogues such as peptoids (see Simon et al., PNAS USA 89 (20): 9367 (1992), incorporated by reference in its entirety). Amino acids may be either natural or synthetic (eg, not the amino acid encoded by DNA); as will be understood by those skilled in the art. For example, for the purposes of this invention, homophenylalanine, citrulline, ornithine, and norleucine are considered synthetic amino acids, and both D- and L-(R or S) amino acid configurations can be used. Options according to this invention may contain modifications that include the use of synthetic amino acids incorporated using, for example, technologies developed by Schultz and colleagues, including but not limited to the methods described in Cropp & Shultz, 2004, Trends Genet. 20(12): 625-30, Anderson et al., 2004, Proc Natl Acad Sci USA 101(2):7566-71, Zhang et al., 2003, 303(5656):371-3, and Chin et al. ., 2003, Science 301(5635):964-7, all incorporated by reference in their entirety. In addition, polypeptides may include synthetic derivatization of one or more side chains or ends, glycosylation, pegylation, cycling, cyclization, linkers to other molecules, fusion to proteins or protein domains, and the addition of peptide tags or markers.

Residue in this document means a position in a protein and its associated amino acid

- 20 040773 identity. For example, asparagine 297 (also referred to as Asn297 or N297) is a residue at position 297 in a human IgG1 antibody.

Fab or Fab region as used herein refers to a polypeptide that contains the VH, CH1, VL and CL immunoglobulin domains. The Fab may refer to this region alone, or to this region in the context of a full length antibody or antibody fragment.

By Fv or Fv fragment or Fv region, as used herein, is meant a polypeptide that contains the VL and VH domains of a single antibody. As will be appreciated by those skilled in the art, they typically consist of two strands.

By single chain Fv or scFv is herein meant a heavy chain variable domain (VH) covalently linked to a light chain variable domain (V _L ), typically using an scFv linker as discussed herein, to form an scFv or scFv domain. . The scFv domain can be in any orientation from N- to C-terminus (V _H -liquor-V _L or V _L -linker-V _H ). In general, the linker is an scFv linker, as is commonly known in the art, the tet residues: technics, the linker peptide predominantly containing the following amino acids Gl·y, Ser, Ala or Thr. The linker peptide must be long enough to link the two molecules so that they assume the correct conformation relative to each other to retain the desired activity. In one embodiment, the linker is about 1 to 50 amino acids in length, preferably 1 to 30 amino acids in length. In one embodiment, linkers from 1 to 20 amino acids in length can be used, with about 5 to about 10 amino acids being used in some embodiments. Suitable linkers include glycine-serine polymers including, for example, (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer, at least one (and in general, usually , 3 to 4), glycine-alanine polymers, alanine-serine polymers, and other flexible linkers. Alternatively, a variety of non-protein polymers may find use as linkers, including but not limited to polyethylene glycol (PEG), polypropylene glycol, polyoxyalkylenes, or copolymers of polyethylene glycol and polypropylene glycol, which may find use as linkers.

By IgG subclass modification or isotype modification, as used herein, is meant an amino acid modification that converts one amino acid of one IgG isotype to the corresponding amino acid of another, aligned IgG isotype. For example, since IgG1 contains tyrosine and IgG2 contains phenylalanine at position EC 296, the F296Y substitution in IgG2 is considered a modification of the IgG subclass. Similarly, since IgG1 has a proline at position 241 and IgG4 has a series, the IgG4 molecule with S241P is considered to be a modification of the IgG subclass. It should be noted that subclass modifications are considered amino acid substitutions herein.

By non-natural modification, as used herein, is meant a modification of an amino acid that is not isotypic. For example, since none of the IgGs contains an asparagine AN at position 297, the substitution of N297A in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a non-natural modification.

Amino acid or amino acid unit in this document means one of the 20 naturally occurring amino acids that are encoded by DNA or RNA.

An effector function, as used herein, means a biochemical event that occurs as a result of interaction between the Fc region of an antibody and an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

An IgG Fc ligand as used herein means a molecule, preferably a polypeptide, derived from any organism that binds to the Fc region of an IgG antibody to form an Fc/Fc ligand complex. Fc ligands include, but are not limited to, FcyRIs, FcyRIIs, FcyRIIIs, FcRn, Clq, C3, mannan-binding lectin, mannose receptor, staphylococcal A protein, streptococcal G protein, and viral FcyR. Fc ligands also include Fc receptor homologues (FcRH), which is a family of Fc receptors that are homologous to FcyR (Davis et al., 2002, Immunological Reviews 190:123136, incorporated by reference in its entirety). Fc ligands may include undiscovered molecules that bind Fc. Specific IgG Fc ligands are FcRn (neonatal Fc receptor) and Fc gamma receptors. An Fc ligand as used herein means a molecule, preferably a polypeptide, derived from any organism that binds to the Fc region of an antibody to form an Fc/Fc ligand complex.

The original polypeptide in this document means the original polypeptide, which is subsequently modified to obtain a variant. The parent polypeptide may be a naturally occurring polypeptide, or a variant or engineered version of a naturally occurring polypeptide. The parent polypeptide may refer to the polypeptide itself, compositions that contain the parent polypeptide, or the amino acid sequence that encodes it. Accordingly, parental immunoglobulin, as used herein, means an unmodified immunoglobulin polypeptide that is modified to produce a variant, and parental antibody, herein, refers to an unmodified antibody that is modified to produce a variant.

- 21 040773 th antibody. It should be noted that the parent antibody includes commercially known, recombinantly produced antibodies, as described below.

Fc, or Fc-region, or Fc-domain herein means a polypeptide containing the constant region of an antibody, with the exception of the first domain of the constant region of the immunoglobulin, and, in some cases, part of the hinge. Thus, Fc refers to the last two domains of the IgA, IgD, and IgG constant regions, the last three domains of the IgE and IgM constant regions, and the N-terminal flexible hinge of these domains. For IgA and IgM, Fc may contain a J chain. For IgG, the Fc domain contains the immunoglobulin domains Cy2 and Cy3 (Cy2 and Cy3) and the lower hinge region between Cy1 (Cy1) and Cy2 (Cy2). Although the boundaries of the Fc region may vary, the Fc region of the heavy chain of human IgG is usually defined as containing residues C226 or P230 at its carboxyl terminus, and the numbering is carried out according to the EC index, as in Kabat. In some embodiments, as described more fully below, amino acid modifications are made in the Fc region, for example, to change binding to one or more FcyR receptors or to an FcRn receptor.

By heavy chain constant region, as used herein, is meant the CH1-hingeCH2-CH3 portion of an antibody.

By position, as used herein, is meant the location in the protein sequence. The positions may be numbered sequentially or according to a fixed format, such as the EC index for antibody numbering.

By target antigen, as used herein, is meant a molecule that is specifically associated with the variable region of a given antibody. Here, one target antigen of interest herein is TIGIT, usually human TIGIT and optionally cynomolgus TIGIT, as defined below. Another target antigen of interest is PVRIG, usually human PVRIG and optionally cynomolgus monkey PVRIG, as defined below.

By target cell, as used herein, is meant a cell that expresses the target antigen.

By variable region, as used herein, is meant a region of an immunoglobulin that contains one or more Ig domains that are essentially encoded by any of the Υκ, Υλ, V _L and/or V _H genes that form the kappa, lambda genetic loci. and immunoglobulin heavy chain, respectively.

Wild type or WT, as used herein, means an amino acid sequence or nucleotide sequence that is found in nature, including allelic variations. The DT protein has an amino acid sequence or a nucleotide sequence that has not been intentionally modified.

Antibodies according to this invention are usually isolated or recombinant. Selected here is used to describe the various polypeptides described herein, refers to a polypeptide that is identified and separated and/or isolated from the cell or cell culture in which it was expressed. Typically, an isolated polypeptide will be obtained using at least one purification step. An isolated antibody refers to an antibody that is substantially free of other antibodies having different antigenic specificities. Recombinant means that antibodies are generated using recombinant nucleic acid techniques in exogenous host cells.

Specific binding, or specifically binds to, or is specific to a particular antigen or epitope, means a binding that is measurably different from a non-specific interaction. Specific binding can be measured, for example, by determining the binding of a molecule compared to the binding of a control molecule, which is usually a molecule with a similar structure that has no binding activity. For example, specific binding may be determined by competition with a control molecule that is similar to the target.

Specific binding for a particular antigen or epitope can be, for example, an antibody having a KD for the antigen or epitope of at least about 10 ^-9 M, at least about 10 ^-10 M, at least about 10 ^-11 M, at least at least about 10 ^-12 M, at least about 10 ^-13 M, at least about 10 ^-14 M, at least about 10 ^-15 M, where KD refers to the rate of dissociation of a particular antibody-antigen interaction. Typically, an antibody that specifically binds an antigen will have a KD that is 20, 50, 100, 500, 1000, 5000, 10,000 or more times the reference molecule relative to the antigen or epitope.

Specific binding of a particular antigen or epitope may also be shown, e.g., an antibody having KA or Ka for the antigen or epitope at least 20, 50, 100, 500, 1000, 5000, 10,000 or more times for the epitope relative to the control, where KA or Ka refers to the rate of association of a particular antibody-antigen interaction. Binding affinity is usually measured

- 22 040773 is measured using surface plasmon resonance (eg Biacore assay) and flow cytometry with antigen-expressing cells.

C. Sequences

The sequence listing contains a number of sequences based on the format of FIG. 53; In FIG. 4 is USSN 62/513916 expressly incorporated by reference herein) as a guideline for labeling sequences. The heavy chain variable domain is labeled with an identifier (eg CPA.0.86) with the following sequence following the format in FIG. 53 of this specification (identical to the format of FIG. 4 mentioned above), in which the next sequence identifier refers to vhCDR1, next to vhCDR2, with vhCDR3, full-length heavy chain, light chain variable domain, VLCDR1, VLCDR2, VLCDR3, and full-length light chain. Thus, a single antibody has 10 associated sequence identifiers.). The list of sequences includes BM26 mouse IgG1 (BM26-M1) (WO 2016/028656A1, clone 31C6) and BM29 mouse IgG1 (BM29-M1) (US 2016/0176963A1, clone 22G2). Unless indicated, full-length HC anti-TIGIT antibody sequences are in H4 (S241P) format.

D. PVRIG proteins

The present invention relates to antibodies that specifically bind to PVRIG proteins and prevent its activation by the PVRL2 ligand protein, a human plasma membrane glycoprotein. PVRIG, also referred to as poliomyelitis virus receptor-associated immunoglobulin domain-containing protein, Q6DKI7 or C7orf15, refers to the amino acid and nucleotide sequences specified in RefSeq attachment identifier NP_076975 depicted in FIG. 1. Sequence of human polio virus receptor-associated protein 2 (PVLR2, also known as nectin-2, CD112, or herpes virus entry mediator B (HVEB), PVRIG binding partner (as shown in Example 5 of US 2016/0244521) shown in Fig. 2. Antibodies according to the invention are specific for the extracellular domain of PVRIG, so that the binding of PVRIG and PVLR2 is blocked.

PVRIG is a 326 amino acid transmembrane domain protein with a signal peptide (ranging from amino acid 1 to 40), an extracellular domain (ranging from amino acid 41 to 171), a transmembrane domain (ranging from amino acid 172 to 190), and a cytoplasmic domain (ranging from amino acid 191). up to 326). There are two methionines that can be start codons, but the mature proteins are identical.

Accordingly, as used herein, the term PVRIG or PVRIG protein or PVRIG polypeptide may optionally include any such protein or variants, conjugates or fragments thereof, including but not limited to known or wild-type PVRIG as described herein, as well as any possible natural splicing variants, amino acid variants or isoforms and, in particular, the PVRIG EVA fragment.

As noted herein and described in more detail below, anti-PVRIG antibodies (including antigen-binding fragments) that both bind to PVRIG and prevent PVRL2 activation (eg, most commonly by blocking the interaction of PVRIG and PVLR2) are used to enhance T activation. α-cells and/or NK cells and can be used to treat diseases such as cancer and pathogenic infection.

E. TIGIT proteins

This invention relates to antibodies that specifically bind to TIGIT proteins and prevent its activation by its ligand protein, PVR, poliovirus receptor (aka CD 155), human plasma membrane glycoprotein. TIGIT or T cell immunoreceptor with Ig and ITIM domains is a co-inhibitory receptor protein also known as WUCAM, Vstm3 or Vsig9. TIGIT has an immunoglobulin variable domain, a transmembrane domain and an immunoreceptor tyrosine inhibitory motif (ITIM) and contains signature sequence elements of the PVR protein family. The extracellular domain (ECD) sequences of TIGIT and PVR are shown in FIG. 51. The antibodies of the invention are specific for TIGIT EVA so that the binding of TIGIT and PVR is blocked.

Accordingly, as used herein, the term TIGIT or TIGIT protein or TIGIT polypeptide may optionally include any such protein or variants, conjugates, or fragments thereof, including but not limited to known or wild-type TIGIT as described herein, as well as any possible natural splicing variants, amino acid variants or isoforms, and in particular the TIGIT EVA fragment.

As noted herein and described in more detail below, anti-TIGIT antibodies (including antigen-binding fragments) that both bind to TIGIT and prevent PVR activation (eg, most commonly by blocking TIGIT-PVR interaction) are used to enhance T activation. α-cells and/or NK cells and can be used to treat diseases such as cancer and pathogenic infection.

- 23 040773

VI. Antibodies

As discussed below, the term antibody is used in a general sense. The structural units of a standard antibody usually contain a tetramer. Each tetramer typically consists of two identical pairs of polypeptide chains, each pair having one light chain (typically having a molecular weight of about 25 kDa) and one heavy chain (typically having a molecular weight of about 50-70 kDa). Human light chains are classified as kappa and lambda light chains. This invention relates to monoclonal antibodies, which are usually based on the IgG class, which has several subclasses, including, but not limited to, IgG1, IgG2, IgG3 and IgG4. In general, IgG1, IgG2 and IgG4 are more commonly used than IgG3. It should be noted that IgG1 has different allotypes with polymorphisms at positions 356 (D or E) and 358 (L or M). The sequences reported herein use the 356D/358M allotype, but a different allotype is included in this document. That is, any sequence containing an IgG1 Fc domain included herein can have 356E/358L replacing the 356D/358M allotype.

The amino terminal portion of each chain contains a variable region of about 100 to 110 or more amino acids primarily responsible for antigen recognition, commonly referred to in the art and herein as the Fv domain or Fv region. Three loops are assembled in the variable region for each of the heavy chain and light chain V domains to form an antigen binding site. Each of the loops is referred to as a complementarity determining region (hereinafter referred to as CDR) in which the amino acid sequence change is most significant. Variable refers to the fact that, among antibodies, certain segments of the variable region differ significantly in sequence. Variability within the variable region is unevenly distributed. In fact, V regions are composed of relatively invariant fragments called framework regions (FRs) of 15-30 amino acids separated by shorter regions of extreme variability called hypervariable regions, each 9-15 amino acids long or longer.

Each VH and VL consists of three hypervariable regions (complementarity determining regions, CDRs) and four FRs, arranged from N-terminus to C-terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

The hypervariable region typically spans amino acid residues from about 24-34 amino acid residues (LCDR1, L stands for light chain), 50-56 (LCDR2) and 89-97 (LCDR3) in the light chain variable region, and about 31-35B (HCDR1; H stands for heavy chain), 50-65 (HCDR2) and 95-102 (HCDR3) in the variable region of the heavy chain; Kabat et al, SEQUENCES OF PROTEINS OF IMMUNOLOGICAL INTEREST, 5 th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991) and/or these residues forming a hypervariable loop (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2) and 91-96 (LCDR3) in the light chain variable region and 26-32 (HCDR1), 53 -55 (HCDR2) and 96101 (HCDR3) in the heavy chain variable region, Chothia and Lesk (1987) J Mol Biol 196:901-917 Specific CDRs of the invention are described below.

As will be appreciated by those skilled in the art, the precise numbering and placement of CDRs may differ among different numbering systems. However, it should be understood that the description of the sequences of the variable region of the heavy chain and/or the variable region of the light chain includes a description of the associated (characteristic) CDR. Accordingly, each heavy chain variable region description is a vhCDR description (eg, vhCDR1, vhCDR2, and vhCDR3), and each light chain variable region description is a vlCDR description (eg, VlCDR1, VlCDR2, and VlCDR3). A useful comparison of CDR numbering is given below, see Lafranc et al., Dev. Associate Immunol. 27( 1):55-77 (2003):_________________________________________________________________

Kabat+Chothia IMGT Kabat AbM Chothia Contact vhCDRl 26-35 27-38 31-35 26-35 26-32 30-35 vhCDR2 50-65 56-65 50-65 50-58 53-55 47-58 vhCDR3 95-102 105-117 95-102 95-102 96-101 93-101 vlCDRl 24-34 27-38 24-34 24-34 26-32 30-36 vlCDR2 50-56 56-65 50-56 50-56 50-52 46-55 vlCDR3 89-97 105-117 89-97 89-97 91-96 89-96

In this specification, the Kabat numbering system is generally used to denote a residue in the variable region (approximately residues 1-107 of the light chain variable region and residues 1113 of the heavy chain variable region), and the hinge and EC numbering system for Fc regions (e.g., Kabat et al., supra (1991)).

The present invention provides a large number of different sets of CDRs. In this case, the complete set of CDRs contains three light chain variable regions and three CDR heavy chain variable regions, eg VlCDR1, VLCDR2, VLCDR3, vhCDR1, vhCDR2 and vhCDR3. They may be part of a larger light chain variable domain or a heavy chain variable domain, respectively. In addition, as described in more detail herein, the heavy and light chain variable domains can be on separate polypeptide chains when heavy and light chains are used, or on the same polypeptide chain in the case of scFv sequences.

- 24 040773

CDRs promote the formation of an antigen-binding, or more specifically, epitope-binding site of antibodies. An epitope refers to a determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule known as a paratope. Epitopes are grouped from molecules, such as amino acids or saccharide side chains, and usually have specific structural characteristics as well as specific charge characteristics. One antigen may have more than one epitope.

An epitope may contain amino acid residues directly involved in binding (also called the immunodominant component of an epitope) and other amino acid residues that are not directly involved in binding, such as amino acid residues that are effectively blocked by a specific antigen-binding peptide; in other words, the amino acid residue is within the recognition zone of the specific antigen-binding peptide.

Epitopes can be either conformational or linear. A conformational epitope is created by spatially aligned amino acids from different segments of a linear polypeptide chain. A linear epitope is an epitope created by adjacent amino acid residues in a polypeptide chain. Conformational and non-conformational epitopes may differ in that in the presence of denaturing solvents, binding to the former but not to the latter epitopes is lost.

An epitope typically contains at least 3, and more typically, at least 5 or 8-10 amino acids in a unique spatial conformation. Antibodies that recognize the same epitope can be tested by a simple immunoassay, such as sorting, to demonstrate the ability of one antibody to block another antibody from binding to the target antigen. As indicated below, the invention includes not only the antigen-binding domains and antibodies listed herein, but also those that compete for binding to epitopes associated with the listed antigen-binding domains.

The carboxy-terminal portion of each chain defines a constant region primarily responsible for effector function. Kabat et al. collected numerous primary sequences of the variable regions of the heavy chains and light chains.

Based on the degree of sequence conservation, they classified individual primary sequences into CDRs and structure and compiled a list of them (see SEQUENCES OF IMMUNOLOGICAL INTEREST, 5th edition, NIH publication, No. 91-3242, E.A. Kabat et al., incorporated by reference in its entirety) .

Within the IgG subclass of immunoglobulins, there are several immunoglobulin domains in the heavy chain. An immunoglobulin (Ig) is herein understood to mean a region of an immunoglobulin having a defined tertiary structure. The subject of interest in this invention are heavy chain domains, including constant heavy (CH) domains and hinge domains. In the context of IgG antibodies, IgG isotypes have three CH regions. Accordingly, the CH domains in the context of IgG are as follows: CH1 refers to positions 118-220 according to the EC index, as in Kabat. CH2 refers to provisions 237-340 according to the EU suffix, as in Kabat, and CH3 refers to provisions 341-447 according to the EC suffix, as in Kabat.

Another type of heavy chain Ig domain is the hinge region. Hinge or hinge region or antibody hinge region or immunoglobulin hinge region as used herein means a flexible polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain ends at position EC 220 and the IgG CH2 domain begins at a residue at position EC 237. Thus, an IgG antibody hinge is defined herein as comprising positions 221 (D221 in IgG1) to 236 (G236 in IgG1), and the numbering corresponds to the EU index, as in Kabat.

The light chain typically consists of two domains, the light chain variable region (containing the light chain CDRs and together with the heavy chain variable domains form the Fv region) and the light chain constant region (often referred to as CL or Ck). In general, a kappa chain constant domain or a lambda chain constant domain can be used, with lambda generally finding use in the invention.

Another area of interest for additional substitutions described below is the Fc region.

A. Chimeric and Humanized Antibodies

In some embodiments, the antibodies herein may be derived from a mixture of different species, such as a chimeric antibody and/or a humanized antibody. In general, both chimeric antibodies and humanized antibodies refer to antibodies that combine regions from more than one species. For example, chimeric antibodies traditionally contain variable(e) region(s) from the mouse (or, in some cases, rats) and constant(s) region(s) from the human body. Humanized antibodies generally refer to non-human antibodies that have the variable domain framework regions replaced with sequences found in human antibodies. Typically, in a humanized antibody, the complete antibody, except for the CDR, is encoded by a polynucleotide of human origin or identical to such

- 25 040773 antibody, except for its CDR. CDRs, some or all of which are encoded by nucleic acids derived from a non-human organism, are grafted onto a human antibody variable region beta sheet scaffold to generate an antibody whose specificity is determined by the grafted CDRs. The production of such antibodies is described, for example, in WO 92/11018, Jones, 1986, Nature 321: 522-525, Verhoeyen et al., 1988, Science 239: 1534-1536, all incorporated by reference in their entirety. Backmutation of selected acceptor framework residues to the corresponding donor residues is often required to restore affinity that is lost in the initial graft construct (US 5530101; US 5585089; US 5693761; US 5693762; US 6180370; US 5859205; US 5821337; US 6054297; US 640721 all incorporated in their entirety by reference). A humanized antibody will also optimally comprise at least a portion, and typically all, of an immunoglobulin constant region, typically a human immunoglobulin, and therefore will typically include a human Fc region. Humanized antibodies can also be generated using mice with genetically modified immune systems. Roque et al., 2004, Biotechnol. Prog. 20: 639-654, incorporated by reference in its entirety. Various techniques and methods for humanizing and modifying non-human antibodies are well known in the art (see Tsurushita & Vasquez, 2004, Humanization of Monoclonal Antibodies, Molecular Biology of B Cells, 533-545, Elsevier Science (USA) and references cited therein. , all incorporated by reference in their entirety). Humanization methods include, but are not limited to, those described in Jones et al., 1986, Nature 321:522-525; Riechmann et al., 1988; Nature 332:323-329; Verhoeyen et al., 1988, Science, 239:1534-1536; Queen et al., 1989, Proc Natl Acad Sci, USA 86:10029-33; He et al., 1998, J. Immunol. 160:1029-1035; Carter et al., 1992, Proc Natl Acad Sci USA 89:4285-9, Presta et al., 1997, Cancer Res. 57(20):4593-9; Gorman et al., 1991, Proc. Natl. Acad. sci. USA 88: 4181-4185; O'Connor et al., 1998, Protein Eng 11: 321-8, all incorporated by reference in their entirety. Humanization or other methods of reducing the immunogenicity of the variable regions of a non-human antibody may include methods of altering the surface, as described, for example, in Roguska et al., 1994, Proc. Natl. Acad. sci. USA 91: 969-973, incorporated by reference in its entirety.

Thus, vhCDR and VLCDR from any of the listed antibodies in this case can be humanized (or rehumanized for those that have already been humanized).

In some embodiments, the antibodies of the invention comprise a heavy chain variable region from a particular germline immunoglobulin heavy chain gene and/or a light chain variable region from a particular germline immunoglobulin light chain gene. For example, such antibodies may contain or consist of a human antibody containing heavy or light chain variable regions that are the product of or derived from a particular germline sequence. A human antibody that is the product of, or derived from, a human germline immunoglobulin sequence can be identified as such by comparing the amino acid sequence of the human antibody with the amino acid sequences of human germline immunoglobulins and selecting the human germline immunoglobulin sequence that is closest in sequence (i.e., the highest percent identity). ) to the human antibody sequence. A human antibody that is the product of or derived from a particular human germline immunoglobulin sequence may contain amino acid differences compared to the germline sequence, for example, due to natural somatic mutations or deliberately introduced site-directed mutation. However, a humanized antibody has at least 90% amino acid sequence identity with the amino acid sequence encoded by the human germline immunoglobulin gene and contains amino acid residues that identify the antibody as derived from human sequences when compared to the amino acid sequences of germline immunoglobulin from other species (e.g., mouse lineage germ sequences). In some instances, a humanized antibody may be at least 95%, 96%, 97%, 98% or 99% or even at least 96%, 97%, 98% or 99% identity in amino acid sequence with the amino acid sequence encoded by the germline immunoglobulin gene, excluding the CDR. Thus, the CDRs may be murine, but the variable region (heavy or light) framework regions may be at least 96%, 97%, 98%, or 99% identical in amino acid sequence to the framework region amino acids encoded by a single human germline immunoglobulin gene.

Typically, a humanized antibody derived from a particular human germline sequence will display no more than 10-20 amino acid differences from the amino acid sequence encoded by the human germline immunoglobulin gene. In some cases, a humanized antibody may show no more than 5, or even no more than 4, 3, 2, or 1 amino acid differences from the amino acid sequence encoded by the germline immunoglobulin gene (again, prior to the introduction of any variant, i.e., the number of variants is usually small ).

In one embodiment, the parent antibody has affinity matured as known in the art. For humanization and maturation of affinity, methods can be applied,

- 26 040773 based on a change in the structure, for example, as described in USSN 11/004590. Selection methods can be used to humanize and/or affinity mature the variable regions of an antibody, including, but not limited to, the methods described in Wu et al, 1999, J. Mol. Biol. 294:151-162; Baca et al., 1997, J. Biol. Chem. 272(16): 10678-10684; Rosok et al., 1996, J. Biol. Chem. 271(37): 22611-22618; Rader et al., 1998, Proc. Natl. Acad. sci. USA 95: 8910-8915; Krauss et al., 2003, Protein Engineering 16(10):753-759, all incorporated by reference in their entirety. Other methods of humanization may include transplantation of only parts of the CDR, including, but not limited to, the methods described in USSN 09/810,510; Tan et al., 2002, J. Immunol. 169:1119-1125; De Pascalis et al., 2002, J. Immunol. 169: 3076-3084, all incorporated by reference in their entirety.

B. Optional Antibody Modifications

Antibodies of the invention may be modified or designed to change amino acid sequences by amino acid substitutions. As discussed herein, amino acid substitutions can be made to change the affinity of the CDR for protein (eg, TIGIT or PVRIG, including both increasing and decreasing binding), as well as to change additional functional properties of antibodies. For example, antibodies can be modified to contain modifications within the Fc region, typically to change one or more functional properties of the antibody, such as serum half-life, complement fixation, Fc receptor binding, and/or antigen-dependent cellular cytotoxicity. In addition, an antibody in accordance with at least some embodiments of the invention may be chemically modified (for example, one or more chemical fragments may be attached to the antibody) or may be modified to change its glycosylation, again to change one or more functional properties of the antibody. . Such embodiments are described below. The numbering of residues in the Fc region refers to the Kabat EC index.

In one embodiment, the C _H1 hinge region is modified such that the number of cysteine residues in the hinge region changes, eg, increases or decreases. This approach is further described in US Pat. No. 5,677,425 to Bodmer et al. The number of cysteine residues in the CH1 hinge region is changed, for example, to facilitate the assembly of light and heavy chains, or to increase or decrease the stability of the antibody.

In yet another embodiment, the antibody can be modified to abolish Fab-arm exchange in vivo, particularly when IgG4 constant domains are used. In particular, this process involves the exchange of IgG4 hemimolecules (one heavy chain plus one light chain) between other IgG4 antibodies, which effectively results in bispecific antibodies that are functionally monovalent. Mutations in the hinge region and constant regions of the heavy chain can suppress this exchange (see Aalberse, RC, Schuurman J., 2002, Immunology 105: 9-19). As stated in this document, the mutation that finds particular use in this invention is S241P in the context of the IgG4 constant domain. IgG4 finds use in the present invention because it has no significant effector function and is therefore used to block receptor binding to its ligand without depleting cells (eg PVRIG with PVRL2 or TIGIT with PVR).

In some embodiments, amino acid substitutions can be made in the Fc region, typically to alter binding to FcγR receptors. The term Fc-gamma receptor, FcyR or FcgammaR, as used herein, means any member of a family of proteins that binds the Fc region of an IgG antibody and is encoded by the FcyR gene. In humans, this family includes, but is not limited to, FcyRI (CD64), including the FcyRIa, FcyRIb, and FcyRIc isoforms; FcyRII (CD32), including FcyRIIa (including H131 and R131 allotypes), FcyRIIb (including FcyRIIb-1 and FcyRIIb-2) and FcyRIIc isoforms; and FcyRIII (CD16), including the isoforms FcyRIIIa (including the V158 and F158 allotypes) and FcyRIIIb (including the FcyRIIIb-NA1 and FcyRIIIb-NA2 allotypes) (Jefferis et al., 2002, Immunol Lett 82: 57-65, incorporated by reference in their entirety) , as well as any undiscovered human FcyRs or FcyR isoforms or allotypes. The FcyR may be from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. Mouse FcyRs include, but are not limited to, FcyRI (CD64), FcyRII (CD32), FcyRIII-1 (CD16), and FcyRIII-2 (CD 16-2), as well as any undiscovered mouse FcyRs or FcyR isoforms or allotypes.

There are a number of useful Fc substitutions that can be made to alter binding to one or more FcyR receptors. Substitutions that result in an increase in binding as well as a decrease in binding may be useful. For example, increased binding to FcyRIIIa is known to generally lead to an increase in ADCC (antibody-dependent cell-mediated cytotoxicity, a cell-mediated reaction in which non-specific cytotoxic cells that express FcyR recognize the bound antibody in the target cell and subsequently cause lysis of the target cell. Similarly, a decrease in some associations with FcyRIIb (an inhibitory receptor) can also be useful in some circumstances.Amino acid substitutions that are used in this invention include those listed in US Sen No. 11/124 620 (in particular, Fig. 41) and US patent No. 6737056, both of which are fully included in this description

- 27 040773 by reference in their entirety and specifically for the options disclosed therein.

In yet another example, the Fc region is modified to increase the ability of the antibody to mediate antibody-dependent cellular cytotoxicity (ADCC) and/or to increase the affinity of the antibody for the Fcy receptor and/or to increase FcRn binding by modifying one or more amino acids at the following positions: 238, 239 , 248, 249, 252, 254, 255, 256, 258, 265, 267, 268, 269, 270, 272, 276, 278, 280, 283, 285, 286, 289, 290, 292, 293, 294, 295 , 296, 298, 301, 303, 305, 307, 309, 312, 315, 320, 322, 324, 326, 327, 329, 330, 331, 333, 334, 335, 337, 338, 340, 360, 373 , 376, 378, 382, 388, 389, 398, 414, 416, 419, 430, 434, 435, 437, 438, or 439. This approach is described further in PCT publication WO 00/42072 from Presta. In addition, binding sites on human IgG1 for Fc1RI, FcyRII, FcyRIII and FcRn have been identified and variants with enhanced binding have been described (See Shields, R. L. et al. (2001) J. Biol. Chem. 276:6591-6604). Specific mutations at positions 256, 290, 298, 333, 334 and 339 have been shown to improve binding to FcyRIII. In addition, the following combination mutants have been shown to improve FcyRIII binding: T256A/S298A, S298A/E333A, S298A/K224A and S298A/E333A/K334A FcRn and increase the half-life of antibodies (see Chan CA and Carter PJ (2010) Nature Rev Immunol 10: 301-316).

In addition, the antibodies according to the invention are modified to increase the biological half-life. Various approaches are possible. For example, you can enter one or more of the following mutations: T252L, T254S, T256F, as described in US patent No. 6277375 to Ward. Alternatively, to increase the biological half-life, the antibody can be modified in the C _H1 or CL region to contain a salvage receptor binding epitope taken from two loops of the CH2 domain of the IgG Fc region, as described in US Patent Nos. 5,869,046 and 6,121,022 to Presta et al. . Additional mutations to increase serum half-life are described in US Pat.

In yet another embodiment, the glycosylation of the antibody is modified. For example, an aglycosylated antibody (i.e., the antibody is devoid of glycosylation) can be produced. Glycosylation can be altered, for example, to increase the affinity of an antibody for an antigen or to decrease an effector function such as ADCC. Such carbohydrate modifications can be made, for example, by changing one or more glycosylation sites within the sequence of an antibody, for example, N297. For example, one or more amino acid substitutions can be made that result in the elimination of one or more variable region framework glycosylation sites, thereby eliminating glycosylation at that site using alanine substitution search in some embodiments.

Additionally or alternatively, an antibody can be produced that has an altered glycosylation pattern, such as a hypofucosylated antibody having reduced amounts of fucosyl residues, or an antibody having increased numbers of GlcNAc branch point structures. These altered glycosylation patterns have been shown to increase the ability of antibodies to ADCC. Such carbohydrate modifications can be accomplished, for example, by expressing an antibody in a host cell with an altered glycosylation mechanism. Cells with altered glycosylation mechanism have been described in the art and can be used as host cells in which recombinant antibodies can be expressed in accordance with at least some embodiments of the invention in order to obtain an antibody with altered glycosylation. See, for example, US Patent Publication No. 20040110704 and WO 2003/035835.

Another modification of the antibodies in this document, which is proposed in the invention, is PEGylation or the addition of other water-soluble fragments, usually polymers, for example, to increase the half-life. An antibody can be PEGylated, for example, to increase the biological (eg, serum) half-life of the antibody, as is known in the art.

In addition to substitutions made to alter binding affinity for FcyRs and/or FcRn and/or to increase in vivo serum half-life, further antibody modifications may be made, as described in more detail below.

In some cases, affinity maturation occurs. Amino acid modifications in CDRs are sometimes referred to as affinity maturation. An affinity matured antibody is an antibody having one or more changes in one or more CDRs that result in an increase in the affinity of the antibody for the antigen compared to the parent antibody not containing such change(s). In some cases, it may be desirable to reduce the affinity of an antibody for its antigen.

In some embodiments, one or more amino acid modifications are made to one or more CDRs of an antibody of the invention (anti-PVRIG or TIGIT antibodies). In general, only 1 or 2 or 3 amino acids are substituted in any single CDR, and usually no more than 1, 2, 3. 4, 5, 6, 7, 8, 9, or 10 changes are made within a set of 6 CDRs (e.g. , vhCDR1-3 and VLCDR1-3). However

- 28 040773 it should be taken into account that any combination of substitutions, 1, 2 or 3 substitutions in any CDR can be independently and optionally combined with any other substitution.

Affinity maturation can be done to increase the binding affinity of the antibody to the antigen by at least about 10% to 50% to 100% to 150% or more, or 1 to 5 times the parent antibody. Preferred affinity matured antibodies will have nanomolar or even picomolar affinity for the antigen. Affinity matured antibodies are prepared using known techniques. The relationship between affinity and potency is discussed below.

Alternatively, amino acid modifications can be made to one or more CDRs of an antibody of the invention that are silent, eg, that do not significantly alter the affinity of the antibody for the antigen. They may be performed for a number of reasons, including optimization of expression (as may be done for nucleic acids encoding antibodies of the invention).

Thus, CDR and antibody variants are included in the definition of CDRs and antibodies according to the invention; that is, the antibodies of the invention may contain amino acid modifications of the amino acids in one or more of the CDRs of the listed antibodies of the invention. In addition, as indicated below, amino acid modifications may also be independently and optionally made in any region outside of the CDR, including the framework and constant regions.

a) Creation of additional antibodies

Additional anti-human PVRIG antibodies can be made as is well known in the art using well known methods such as those described in the examples. Thus, additional anti-PVRIG antibodies can be generated by standard methods such as immunizing mice (sometimes using DNA immunization, such as using Aldevron), followed by screening against human PVRIG protein and hybridoma generation with antibody purification and recovery.

VII. Antibodies to TIGIT according to the invention

The present invention provides anti-TIGIT antibodies (for convenience, anti-TIGIT antibodies and anti-TIGIT antibodies are used interchangeably). The anti-TIGIT antibodies of the invention bind specifically to human TIGIT and preferably to human TIGIT EVA. The invention further provides antigen binding domains, including full length antibodies, which contain a number of specific enumerated sets of 6 CDRs that bind to TIGIT.

Specific binding for TIGIT or a TIGIT epitope can be detected, for example, with an antibody having a KD of at least about ^10-4 M, at least about ^10-5 M, at least about 10-6 M, at least about 10 - ⁷ M, at least about ^10-8 M, at least about ^10-9 M, at least about ^10-10 M, at least about ^10-11 M, at least about ^10-12 M, according to at least about ^10-13 M, at least about ^10-14 M, at least about ^10-15 M, or greater, wherein KD refers to the equilibrium dissociation constant of a particular antibody-antigen interaction. Typically, an antibody that specifically binds an antigen will have a KD that is 20, 50, 100, 500, 1000, 5000, 10,000 or more times greater for a control molecule compared to the TIGIT antigen or epitome.

However, for optimal binding to TIGIT expressed on the surface of NK and T cells, antibodies preferably have a KD of less than 50 nM and most preferably less than 1 nM, when less than 0.1 nM and less than 1 pM are used in the methods of this invention.

In addition, specific binding for a particular antigen or epitope can be detected, for example, with an antibody having a ka (in reference to the association rate constant) for the TIGIT antigen or epitope of at least 20, 50, 100, 500, 1000, 5000, 10,000 or more times for an epitope compared to a control, where ka refers to the association rate constant of a particular antibody-antigen interaction.

In some embodiments, anti-TIGIT antibodies of the invention bind to human TIGIT with a KD of 100 nM or less, 50 nM or less, 10 nM or less, or 1 nM or less (i.e., higher binding affinity), or 1 pM, or less, with KD determined by known methods, for example, surface plasmon resonance (SPR, for example, Biacore assays), ELISA (enzyme-linked immunosorbent assay), KINEXA (kinetic exclusion assay) and most typically SPR at 25 or 37°C.

Anti-TIGIT antibodies described herein are labeled as follows. Antibodies have codes, for example, CPA.9.086. They display a combination of heavy and variable light chain variable regions as depicted in FIG. 53, for example, with the understanding that these antibodies contain two heavy chains and two light chains. CPA.9.086.VH refers to the variable portion of the CPA heavy chain. 9.086, while CPA. 9.086.VL is the variable region of the light chain. SRA. 9.086.vhCDR1, CPA. 9.086.vhCDR2, CPA. 9.086.vhCDR3, CPA. 9.086.VLCDR1, CPA. 9.086.VLCDR2 and CPA. 9.086.VLCDR3, refers to CDR. SRA. 9.086.HC refers to the entire heavy chain (eg, variable and constant domain) of that molecule, and CPA.9.086.LC refers to the entire light chain (eg, variable and constant domain) of the same molecule. In general, easy

- 29 040773 The human kappa chain is used for the constant region of each phage antibody (or humanized hybridoma) herein, although in some embodiments, the lambda light chain constant domain is used. SRA. 9.086.H1 refers to a full-length antibody containing heavy and light chain variable domains, including the human (Human) IgG1 constant domain (hence H1; the sequences of IgG1, IgG2, IgG3 and IgG4 are depicted in Fig. 50). According to SRA. 9.086.H2 will be CPA variable domains. 9.086 associated with human (Human) IgG2. SRA. 9.086.H3 will be the CPA variable domains. 9.086 associated with human (Human) IgG3, and CPA. 9.086.H4 will be the CPA variable domains. 9.086 associated with human (Human) IgG4. Note that in some cases human IgGs may have additional mutations, which are described below and this will be annotated. For example, in many embodiments, there may be an S241P mutation in human IgG4, and this may, for example, be annotated as CPA.9.086.H4 (S241P). The sequence of human IgG4 with this S241P hinged variant is depicted in FIG. 50. Other possible variants are IgG1 (N297A) (or other variants that lack glycosylation at this site and hence many of the effector functions associated with FcYRIIIa binding) and IgG1 (D265A), which reduces binding to FcyR receptors.

The invention also provides heavy and light chain variable domains as well as full length heavy and light chains.

In some embodiments, the invention provides scFvs that bind to a TIGIT comprising a heavy chain variable domain and a light chain variable domain linked by an scFv linker as described above. The VL and VH domains can be in any orientation, such as N- to C-terminus of VH-linker-VL or VL-linker-VH. They are named after their component parts; e.g. scFv-CPA. 9.086.VH-linker-VL or scFv-CPA.9.086.VL-linker-VH. Thus, scFvCPA.9.086 can be in any orientation.

In many embodiments, the antibodies of the invention are human (derived from phage) and block the binding of TIGIT and PVR. As shown in FIG. 58 and 75, CPA antibodies that both bind and block the receptor-ligand interaction as follows, with their components listed (as described in the Sequence section, the sequences of all but the scFv constructs are in the sequence listing):

CPA.9.018, CPA.9.018.VH, CPA.9.018.VL, CPA.9.018.HC, CPA.9.018.LC,

CPA.9.018.H1, CPA.9.018.H2, CPA.9.018.N3, CPA.9.018.H4; CPA.9.018.H4(S241P);

CPA.9.018.vhCDRl, CPA.9.018.vhCDR2, CPA.9.018.vhCDR3, CPA.9.018.vlCDRl,

CPA.9.018.V1CDR2, CPA.9.018.vlCDR3 and scFv-CPA.9.018;

CPA.9.027, CPA.9.027.VH, CPA.9.027.VL, CPA.9.027.HC, CPA.9.027.LC,

CPA.9.027.H1, CPA.9.027.H2, CPA.9.027.H3, CPA.9.027.H4; CPA.9.018.H4(S241P);

CPA.9.027.vhCDRl, CPA.9.027.vhCDR2, CPA.9.027.vhCDR3, CPA.9.027.vlCDRl,

CPA.9.027.V1CDR2, CPA.9.027.vlCDR3 and scFv-CPA.9.027;

CPA.9.049, CPA.9.049.VH, CPA.9.049.VL, CPA.9.049.HC, CPA.9.049.LC,

CPA.9.049.H1, CPA.9.049.H2, CPA.9.049.H3; CPA.9.049.H4; CPA.9.049.H4(S241P);

CPA.9.049.vhCDRl, CPA.9.049.vhCDR2, CPA.9.049.vhCDR3, CPA.9.049.vlCDRl,

CPA.9.049.V1CDR2, CPA.9.049.vlCDR3 and scFv-CPA.9.049;

CPA.9.057, CPA.9.057.VH, CPA.9.057.VL, CPA.9.057.HC, CPA.9.057.LC,

CPA.9.057.H1, CPA.9.057.H2, CPA.9.057.H3; CPA.9.057.H4; CPA.9.057.H4(S241P);

- 30 040773

CPA.9.057.vhCDRl, CPA.9.057.vhCDR2, CPA.9.057.vhCDR3, CPA.9.057.vlCDRl,

CPA.9.057.vlCDR2, CPA.9.057.vlCDR3 and scFv-CPA.9.057;

CPA.9.059, CPA9.059.VH, CPA.9.059.VL, CPA.9.059.HC, CPA.9.059.LC,

CPA.9.059.H1, CPA.9.059.H2, CPA.9.059.H3; CPA.9.059.H4; CPA.9.059.H4(S241P);

CPA.9.059.vhCDRl, CPA.9.059.vhCDR2, CPA.9.059.vhCDR3, CPA.9.059.vlCDRl,

CPA.9.059.V1CDR2, CPA.9.059.vlCDR3 and scFv-CPA.9.059;

CPA.9.083, CPA.9.083.VH, CPA.9.083.VL, CPA.9.083.HC, CPA.9.083.LC,

CPA.9.083.H1, CPA.9.083.H2, CPA.9.083.H3; CPA.9.083.H4; CPA.9.083.H4(S241P);

CPA.9.083.vhCDRl, CPA.9.083.vhCDR2, CPA.9.083.vhCDR3, CPA.9.083.vlCDRl,

CPA.9.083.vlCDR2, CPA.9.083.vlCDR3 and scFv-CPA.9.083;

CPA.9.086, CPA.9.086.VH, CPA.9.086.VL, CPA.9.086.HC, CPA.9.086.LC,

CPA.9.086.H1, CPA.9.086.H2, CPA.9.086.H3; CPA.9.086.H4; CPA.9.086.H4(S241P);

CPA.9.086.vhCDRl, CPA.9.086.vhCDR2, CPA.9.086.vhCDR3, CPA.9.086.vlCDRl,

CPA.9.086.vlCDR2, CPA.9.086.vlCDR3 and scFv-CPA.9.086;

CPA.9.089, CPA.9.089.VH, CPA.9.089.VL, CPA.9.089.HC, CPA.9.089.LC,

CPA.9.089.H1, CPA.9.089.H2, CPA.9.089.H3; CPA.9.089.H4; CPA.9.089.H4(S241P);

CPA.9.089.vhCDRl, CPA.9.089.vhCDR2, CPA.9.089.vhCDR3, CPA.9.089.vlCDRl,

CPA.9.089.vlCDR2, CPA.9.089.vlCDR3 and scFv-CPA.9.089;

CPA.9.093, CPA.9.093.VH, CPA.9.093.VL, CPA.9.093.HC, CPA.9.093.LC,

CPA.9.093.H1, CPA.9.093.H2, CPA.9.093.H3; CPA.9.093.H4; CPA.9.093.H4(S241P);

CPA.9.093.vhCDRl, CPA.9.093.vhCDR2, CPA.9.093.vhCDR3, CPA.9.093.vlCDRl,

CPA.9.093.vlCDR2, CPA.9.093.vlCDR3 and scFv-CPA.9.093;

CPA.9.101, CPA.9.101.VH, CPA9.101.VL, CPA.9.101.HC, CPA.9.101.LC,

CPA.9.101.H1, CPA.9.101.H2, CPA.9.101.H3; CPA.9.101.H4; CPA.9.101.H4(S241P);

CPA.9.101.vhCDRl, CPA.9.101.vhCDR2, CPA.9.101.vhCDR3, CPA.9.101.vlCDRl,

CPA.9.101.V1CDR2, CPA.9.101.vlCDR3 and scFv-CPA.9.101; And

CPA.9.103, CPA.9.103.VH, CPA.9.103.VL, CPA.9.103.HC, CPA.9.103.LC,

CPA.9.103.Hl, CPA.9.103.H2, CPA.9.103.H3; CPA.9.103.H4; CPA.9.103.H4(S241P);

CPA.9.103.vhCDRl, CPA.9.103.vhCDR2, CPA.9.103.vhCDR3, CPA.9.103.vlCDRl,

CPA.9.103.V1CDR2, CPA.9.103.vlCDR3 and scFv-CPA.9.103.

In addition, this invention relates to a number of CHA antibodies, which are mouse antibodies derived from hybridomas. As is well known, the six CDRs are useful when placed in either human heavy chain variable region and light chain variable region framework, or when the heavy chain variable domains and light chain variable domains are humanized.

Accordingly, this invention relates to antibodies, usually full-length or scFv domains, which contain the following sets of CDRs, the sequences of which are depicted in FIG. 53 and/or in the sequence listing:

- 31 040773

CHA.9.536.1, CHA9.536.1 VH, CHA.9.536.1. VL, CHA.9.536.1.HC, CHA.9.536. l.LC, CHA.9.536.1.Hl, CHA.9.536.1. H2, CHA.9.536.1. NZ; CHA.9.536.1.H4,

CHA.9.536.1.H4(S241P), CHA.9.536.1.vhCDR1, CHA.9.536.1.vhCDR2,

CHA.9.536.1.vhCDR3, CHA.9.536.1.vlCDRl, CHA.9.536.1.vlCDR2 and CHA.9.536.1.vhCDR3;

CHA.9.536.3, CHA.9.536.3.VH, CHA.9.536.3.VL, CHA.9.536.3.HC, CHA.9.536.3.LC,

CHA.9.536.3.H1, CHA.9.536.3.H2, CHA.9.536.3.H3; CHA.9.536.3.H4,

CHA.9.536.3.H4(S241P); CHA.9.536.3.vhCDR1, CHA.9.536.3.vhCDR2,

CHA.9.536.3.vhCDR3, CHA.9.536.3.vlCDRl, CHA.9.536.3.vlCDR2 and CHA.9.536.3.vhCDR3;

CHA.9.536.4, CHA.9.536.4.VH, CHA.9.536.4.VL, CHA.9.536.4.HC, CHA.9.536.4.LC,

CHA.9.536.4.H1, CHA.9.536.4.H2, CHA.9.536.4.H3; CHA.9.536.4.H4,

CHA.9.536.4.H4(S241P), CHA.9.536.4.vhCDRl, CHA.9.536.4.vhCDR2,

CHA.9.536.4.vhCDR3, CHA.9.536.4.vlCDRl, CHA.9.536.4.vlCDR2 and CHA.9.536.4.vhCDR3;

CHA.9.536.5, CHA.9.536.5.VH, CHA.9.536.5.VL, CHA.9.536.5.HC, CHA.9.536.5.LC,

CHA.9.536.5.H1, CHA.9.536.5.H2, CHA.9.536.5.H3; CHA.9.536.5.H4,

CHA.9.536.5.H4(S241P), CHA.9.536.5.vhCDRl, CHA.9.536.5.vhCDR2,

CHA.9.536.5.vhCDR3, CHA.9.536.5.vlCDRl, CHA.9.536.5.vlCDR2 and CHA.9.536.5.vhCDR3;

CHA.9.536.6, CHA.9.536.6.VH, CHA.9.536.6.VL, CHA.9.536.6.HC, CHA.9.536.6.LC, CHA.9.536.6.H1, CHA.9.536. 6.H2, CHA.9.536.6.H3; CHA.9.536.6.H4, CHA.9.536.6.vhCDRl, CHA.9.536.6.vhCDR2, CHA.9.536.6.vhCDR3, CHA.9.536.6.vlCDRl, CHA.9.536.6.vlCDR2 and CHA. 9.536.6.vhCDR3;

CHA.9.536.7, CHA.9.536.7.VH, CHA.9.536.7.VL, CHA.9.536.7.HC, CHA.9.536.7.LC,

CHA.9.536.7.H1, CHA.9.536.7.H2, CHA.9.536.7.H3; CHA.9.536.7.H4,

CHA.9.536.5.H4(S241P); CHA.9.536.7.vhCDRl, CHA.9.536.7.vhCDR2,

CHA.9.536.7.vhCDR3, CHA.9.536.7.vlCDRl, CHA.9.536.7.vlCDR2 and CHA.9.536.7.vhCDR3;

CHA.9.536.8, CHA.9.536.8.VH, CHA.9.536.8.VL, CHA.9.536.8.HC, CHA.9.536.8.LC,

CHA.9.536.8.H1, CHA.9.536.8.H2, CHA.9.536.8.H3; CHA.9.536.8.H4,

CHA.9.536.8.H4(S241P), CHA.9.536.8.vhCDR1, CHA.9.536.8.vhCDR2,

CHA.9.536.8.vhCDR3, CHA.9.536.8.vlCDRl, CHA.9.536.8.vlCDR2 and CHA.9.536.8.vhCDR3;

CHA.9.560.1, CHA.9.560.1 9.560.1VH, CHA. 9.560.1.VL, CHA. 9.560.1.HC, CHA. 9.560.1.LC, CHA. 9.560.1.H1, CHA. 9.560.1.H2, CHA. 9.560.1.H3; SNA. 9.560.1.H4, CHA.

- 32 040773

9.560. l.H4(S24 IP), CHA. 9.560.l.vhCDRl, CHA. 9.560.l.vhCDR2, CHA. 9.560. l.vhCDR3, CHA. 9.560.l.vlCDRl, CHA. 9.560. l.vlCDR2 and CHA. 9.560.l.vhCDR3;

CHA.9.560.3, CHA.9.560.3 9.560. 3VH, CHA. 9.560. 3.VL, CHA. 9.560. 3.HC, CHA. 9.560. 3.LC, CHA. 9.560. 3.H1, CHA. 9.560. 3.H2, CHA. 9.560. 3.H3; CHA.9.560.3.H4, CHA.9.560.3.H4(S241P); CHA. 9.560. 3.vhCDR1, CHA. 9.560. 3.vhCDR2, CHA. 9.560. 3.vhCDR3, CHA. 9.560. 3.vlCDRl, CHA. 9.560. 3.vlCDR2 and CHA. 9.560. 3.vhCDR3;

CHA.9.560.4, CHA.9.560.4 9.560. 4VH, CHA. 9.560. 4.VL, CHA. 9.560. 4.HC, CHA. 9.560. 4.LC, CHA. 9.560. 4.H1, CHA. 9.560. 4.H2, CHA. 9.560. 4.H3; CHA.9.560.4.H4, CHA.9.560.4.H4 (S241P), CHA. 9.560. 4.vhCDR1, CHA. 9.560. 4.vhCDR2, CHA. 9.560. 4.vhCDR3, CHA. 9.560. 4.vlCDRl, CHA. 9.560. 4.vlCDR2 and CHA. 9.560. 4.vhCDR3;

CHA.9.560.5, CHA.9.560.5 9.560. 5VH, CHA. 9.560. 5.VL, CHA. 9.560. 5.HC, CHA. 9.560. 5.LC, CHA. 9.560. 5.Hl, CHA. 9.560. 5.H2, CHA. 9.560. 5.H3; CHA. 9.560. 5.H4, CHA. 9.560. 5.vhCDR1, CHA. 9.560. 5.vhCDR2, CHA. 9.560. 5.vhCDR3, CHA. 9.560. 5.vlCDRl, CHA. 9.560. 5.V1CDR2 and CHA. 9.560. 5.vhCDR3;

CHA.9.560.6, CHA.9.560.6 9.560. 6VH, CHA. 9.560. 6.VL, CHA. 9.560. 6.HC, CHA. 9.560. 6.LC, CHA. 9.560. 6.H1, CHA. 9.560. 6.H2, CHA. 9.560. 6.H3; CHA.9.560.6.H4, CHA.9.560.6.H4 (S241P), CHA. 9.560. 6.vhCDR1, CHA. 9.560. 6.vhCDR2, CHA. 9.560. 6.vhCDR3, CHA. 9.560. 6.vlCDRl, CHA. 9.560. 6.vlCDR2 and CHA. 9.560. 6.vhCDR3;

CHA.9.560.7, CHA.9.560.7 9.560. 7VH, CHA. 9.560. 7.VL, CHA. 9.560. 7.HC, CHA. 9.560. 7.LC, CHA. 9.560. 7.H1, CHA. 9.560. 7.H2, CHA. 9.560. 7.H3; CHA.9.560.7.H4; CHA.9.560.7.H4 (S241P); CHA. 9.560. 7.vhCDR1, CHA. 9.560. 7.vhCDR2, CHA. 9.560. 7.vhCDR3, CHA. 9.560. 7.vlCDRl, CHA. 9.560. 7.vlCDR2 and CHA. 9.560. 7.vhCDR3;

CHA.9.560.8, CHA.9.560.8 9.560. 8VH, CHA. 9.560. 8.VL, CHA. 9.560. 8.HC, CHA. 9.560. 8.LC, CHA. 9.560. 8.H1, CHA. 9.560. 8.H2, CHA. 9.560. 8.H3; CHA.9.560.8.H4, CHA.9.560.8.H4 (S241P); CHA. 9.560. 8.vhCDR1, CHA. 9.560. 8.vhCDR2, CHA. 9.560. 8.vhCDR3, CHA. 9.560. 8.vlCDRl, CHA. 9.560. 8.vlCDR2 and CHA. 9.560. 8.vhCDR3;

CHA.9.546.1, CHA.9.546.1 9.546.1VH, CHA. 9.546.1.VL, CHA. 9.546.1.HC, CHA. 9.546.1.LC, CHA. 9.546.1.H1, CHA. 9.546.1.H2, CHA. 9.546.1.H3; CHA.9.546.1.H4, CHA.9.546.1.H4 (S241P), CHA. 9. 546.l.vhCDRl, CHA. 9.546.1.vhCDR2, CHA. 9.546.1.vhCDR3, CHA. 9.546.1.vlCDRl, CHA. 9.546.1.vlCDR2 and CHA. 9.546.1.vhCDR3;

CHA.9.547.1, CHA.9.547.1 9.547.1VH, CHA. 9.547.1.VL, CHA. 9.547.1.HC, CHA. 9.547.1.LC, CHA. 9.547.1.H1, CHA. 9.547.1.H2, CHA. 9.547.1.H3; CHA.9.547.1.H4, CHA.9.547.1.H4 (S241P), CHA. 9. 547.l.vhCDRl, CHA. 9.547.1.vhCDR2, CHA. 9.547.1.vhCDR3, CHA. 9.547.1.vlCDRl, CHA. 9.547.1.vlCDR2 and CHA. 9.547.1.vhCDR3;

CHA.9.547.2, CHA. 9.547.2VH, CHA. 9.547.2.VL, CHA. 9.547.2.HC, CHA. 9.547.

- 33 040773

2.LC, CHA. 9.547.2.H1, SHA. 9.547.2.H2, CHA. 9.547.2.NZ; SHA.9.547.2.H4, SHA.9.547.2.H4 (S241P), SHA. 9.547.2.vhCDRl, CHA. 9.547.2.vhCDR2, CHA. 9.547.2.vhCDR3, CHA. 9.547.2.V1CDR1, CHA. 9.547.2.V1CDR2 and CHA. 9.547.2.vhCDR3;

SHA.9.547.3, SHA. 9.547.3VH, SHA. 9.547.3.VL, CHA. 9. 547. Z.NS, SNA. 9.547.3.LC, CHA. 9. 547. Z.N1, SNA. 9. 547. Z.N2, SNA. 9.547.Z.NZ; CHA.9.547.3.H4, CHA.9.547.3.H4 (S241P), CHA. 9.547.3.vhCDRl, CHA. 9.547. 3.vhCDR2, SHA. 9.547.3.vhCDR3, CHA. 9.547.3.V1CDR1, CHA. 9.547.3.V1CDR2 and CHA. 9.547.3.vhCDR3;

SHA.9.547.4, SHA. 9.547.4VH, SHA. 9.547.4.VL, CHA. 9.547.4.NS, SNA. 9.547. 4.LC, SNA. 9.547.4.H1, SHA. 9.547.4.H2, SHA. 9.547.4.NZ; SHA.9.547.4.H4, SHA.9.547.4.H4 (S241P), SHA. 9.547.4.vhCDRl, CHA. 9.547.4.vhCDR2, CHA. 9.547.4.vhCDR3, CHA. 9.547.4.V1CDR1, CHA. 9.547.4.V1CDR2 and CHA. 9.547.4.vhCDR3;

SHA.9.547.6, SHA. 9.547.6 VH, CHA. 9.547.6.VL, CHA. 9.547.6.NS, SNA. 9.547.6.LC, CHA. 9.547.6.H1, SNA. 9.547.6.H2, SHA. 9.547.6.NZ; SHA.9.547.6.H4, SHA.9.547.6.H4 (S241P), SHA. 9.547.6.vhCDRl, CHA. 9.547.6.vhCDR2, CHA. 9.547.6.vhCDR3, CHA. 9.547.6.V1CDR1, CHA. 9.547.6.V1CDR2 and CHA. 9.547.6.vhCDR3;

SHA.9.547.7, SHA. 9.547.7VH, SHA. 9.547.7.VL, CHA. 9.547.7.NS, SNA. 9.547.7.LC, CHA. 9.547.7.H1, SNA. 9.547.7.H2, SHA. 9.547.7.NZ; SHA.9.547.7.H4, SHA.9.547.7.H4 (S241P), SHA. 9.547.7.vhCDRl, CHA. 9.547.7.vhCDR2, CHA. 9.547.7.vhCDR3, CHA. 9.547.7.V1CDR1, CHA. 9.547.7.V1CDR2 and CHA. 9.547.7.vhCDR3;

SHA.9.547.8, SHA. 9.547.8VH, SHA. 9.547.8.VL, CHA. 9.547.8.HC, CHA.9.547.8.LC, CHA. 9.547.8.H1, SNA. 9.547.8.H2, SHA. 9.547.8.NZ; SHA.9.547.8.H4, SHA.9.547.8.H4 (S241P), SHA. 9.547.8.vhCDRl, CHA. 9.547.8.vhCDR2, CHA. 9.547.8.vhCDR3, CHA. 9.547.8.V1CDR1, CHA. 9.547.8.V1CDR2 and CHA. 9.547.8.vhCDR3;

CHA.9.547.9, CHA.9.547.9, CHA.9.547.9VH, CHA.9.547.9.VL, CHA.9. 547.9.HC, CHA.9.547.9.LC, CHA.9.547.9.H1, CHA.9.547.9.H2, CHA.9.547.9.H3; CHA.9.547.9.H4, CHA.9.547.9.H4, CHA.9.547.9.H4(S241P), CHA.9.547.9.H4(S241P), CHA.9.547.9.vhCDRl, CHA.9.547. 9.vhCDR2, CHA.9.547.9.vhCDR3, CHA.9.547.9.vlCDRl, CHA.9.547.9.vlCDR2 and CHA.9.547.9.vhCDR3;

CHA.9.547.13, CHA.9.547.13, CHA.9.547. 13VH, CHA.9. 547.13.VL, CHA.9. 547.13.HC, CHA. 9.547.13.LC, CHA. 9.547.13.H1, CHA.9.547.13.H2, CHA.9. 547.13.H3; CHA.9.547.13.H4, CHA.9.547.13.H4, CHA.9.547.13.H4(S241P), CHA.9.547.13.H4(S241P), CHA. 9.547.13.vhCDR1, CHA.9.547.13.vhCDR2, CHA.9.547. 13.vhCDR3, CHA. 9.547.13.V1CDR1, CHA. 9.547.13.vlCDR2 and CHA. 9.547.13.vhCDR3;

CHA.9.541.1, CHA. 9.541.1.VH, CHA. 9.541.1.VL, CHA. 9.541.1.HC, CHA. 9.541.1.LC, CHA. 9.541.1.H1, CHA. 9.541.1.H2, CHA. 9.541.1.H3; CHA.9.541.1.H4,

- 34 040773

CHA.9.541.1.H4(S241P), CHA. 9. 541.1.vhCDRl, CHA. 9.541.1.vhCDR2, CHA. 9.

41.1.vhCDR3, CHA. 9. 541.1.vlCDRl, CHA. 9. 541.1.vlCDR2 and CHA. 9.541. l.vhCDR3;

SHA.9.541.3, SHA. 9.541.3.VH, SHA. 9.541.3.VL, CHA. 9. 541. Z.NS, SNA. 9.541.

.LC, CHA. 9. 541. Z.N1, SNA. 9. 541. Z.N2, SNA. 9.541.Z.NZ; CHA.9.541.3.H4, CHA.9.541.3.H4 (S241P), CHA. 9.541.3.vhCDRl, CHA. 9.541.3.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.3.V1CDR1, CHA. 9.541.3.V1CDR2 and CHA. 9.541. 3.vhCDR3;

SHA.9.541.4, SHA. 9. 541.4.VH, SHA. 9.541.4.VL, SHA. 9.541.4.NS, SNA. 9.541.

.LC, CHA. 9.541.4.H1, SNA. 9.541.4.H2, SHA. 9.541.4.NZ; SHA.9.541.4.H4, SHA.9.541.4.H4 (S241P), SHA. 9.541.4.vhCDRl, CHA. 9.541.4.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.4.V1CDR1, CHA. 9.541.4.V1CDR2 and CHA. 9.541. 4.vhCDR3;

CHA9.541.5, CHA. 9.541.5.VH, SNA. 9.541.5.VL, CHA. 9.541.5.NS, SNA. 9.541.

.LC, CHA. 9.541.5.H1, SNA. 9.541.5.H2, SHA. 9.541.5.NZ; SHA.9.541.5.H4, SHA.9.541.5.H4 (S241P), SHA. 9.541.5.vhCDRl, CHA. 9.541.5.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.5.V1CDR1, CHA. 9.541.5.V1CDR2 and CHA. 9.541. 5.vhCDR3;

SHA.9.541.6, SHA. 9.541.6.VH, SHA. 9.541.6.VL, CHA. 9.541.6.NS, SNA. 9.541.

.LC, CHA. 9.541.6.H1, SNA. 9.541.6.H2, SHA. 9.541.6.NZ; SHA.9.541.6.H4, SHA.9.541.6.H4 (S241P), SHA. 9.541.6.vhCDRl, CHA. 9.541.6.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.6.V1CDR1, CHA. 9.541.6.V1CDR2 and CHA. 9.541. 6.vhCDR3;

SHA.9.541.7, SHA. 9.541.7.VH, SNA. 9.541.7.VL, SHA. 9.541.7.NS, SNA. 9.541.

.LC, CHA. 9.541.7.H1, SNA. 9.541.7.H2, SHA. 9.541.7.NZ; SHA.9.541.7.H4, SHA.9.541.7.H4 (S241P), SHA. 9.541.7.vhCDRl, CHA. 9.541.7.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.7.V1CDR1, CHA. 9.541.7.V1CDR2 and CHA. 9.541. 7.vhCDR3; And

SNA.9.541.8, SNA. 9.541.8.VH, SNA. 9.541.8.VL, SHA. 9.541.8.NS, SNA. 9.541.

.LC, CHA. 9.541.8.H1, SNA. 9.541.8.H2, SHA. 9.541.8.NZ; SHA.9.541.8.H4, SHA.9.541.8.H4 (S241P); SNA. 9.541.8vhCDRl, CHA. 9.541.8.vhCDR2, CHA. 9.541.

.vhCDR3, SHA. 9.541.8.V1CDR1, CHA. 9.541.8.V1CDR2 and CHA. 9.541. 8.vhCDR3.

In the case of scFvs containing antibody CDRs above, they are designated scFvs, which include an scFv containing a heavy chain variable domain with vhCDR, a linker, and a light chain variable domain with VLCDR, again as above, in any orientation.

Thus the invention includes scFv-CHA.9.536.3.1, scFv-CHA.9.5 36.3, scFv-CHA.9.536.4, scFv-CHA.9.536.5, scFv-CHA.9.536.7, scFv-CHA.9.536.8, scFv -CHA.9.560.1, scFv-CHA.9.560.3, scFv-CHA.9.560.4, scFv-CHA.9.560.5, scFv-CHA.9.560.6, scFvCHA.9.560.7, scFv-CHA.9.560 .8, scFv-CHA.9.546.1, scFv-CHA.9.547.1, scFv-CHA.9.547.2, scFv-CHA.9.547.3, scFv-CHA.9.547.4, scFv-CHA.9.547.6 , scFv-CHA.9.547.7, scFvCHA.9.547.8, scFv-CHA.9.547.9, scFv-CHA.9.547.13, scFv-CHA.9.541.1, scFv-CHA.9.541.3, scFv-CHA .9.541.4, scFv-CHA.9.541.5, scFv-CHA.9.541.6, scFv-CHA.9.541.7 and scFvCHA.9.541.8.

In addition, CHA.9.543 binds to TIGIT but does not block TIGIT-PVR interaction.

As discussed herein, the invention further provides variants of the above components (CPA and CHA), including variants in the CDRs as described above. Thus, the invention provides antibodies containing a set of 6 CDRs as described herein, which may contain one, two or three amino acid differences in the CDR set if the antibody still binds to TIGIT. Appropriate assays for checking if an anti-TIGIT antibody is one that contains mutations compared to the CDR sequences described herein are known in the art, such as Biacore assays.

In addition, the invention further provides variants of the above heavy and light chain variable regions. In this case, the heavy chain variable regions can be 80%, 90%, 95%, 98%, or 99% identical to the VH sequences herein and/or contain from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 amino acid changes or more when Fc variants are used. Light chain variable regions are provided that may be 80%, 90%, 95%, 98% or 99% identical to the VL sequences herein (and in particular CPA.9.086) and/or contain from 1, 2, 3, 4, 5 , 6, 7, 8, 9, 10 or more amino acid changes when Fc variants are used. In these embodiments, the invention includes these options if the antibody still binds to TIGIT. Appropriate assays for checking if an anti-TIGIT antibody is one that contains mutations compared to the CDR sequences described herein are known in the art, such as Biacore assays.

- 35 040773

In addition, heavy and light chains are provided that are 80%, 90%, 95%, 98% or 99% identical to the full length HC and LC sequences herein (and in particular CPA.9.086) and/or contain from 1, 2, 3 , 4, 5, 6, 7, 8, 9, 10 or more amino acid changes when Fc variants are used. In these embodiments, the invention includes these options if the antibody still binds to TIGIT. Appropriate assays for checking if an anti-TIGIT antibody is one that contains mutations compared to the CDR sequences described herein are known in the art, such as Biacore assays.

In addition, the framework regions of heavy and variable light chain variable regions or CPA or CHA antibodies herein can be humanized (or, in the case of CHA antibodies, rehumanized, to the extent that alternative humanization methods can be done), as is known in of the art (with random variants generated in the CDR as needed), and thus humanized variants of the VH and VL circuits of FIG. 53 (and in particular CPA.9.086). In addition, the humanized heavy and light chain variable domains can then be fused to human constant regions, such as constant regions from IgG1, IgG2, IgG3 and IgG4 (including IgG4 (S241P)).

In particular, as known in the art, mouse VH and VL chains can be humanized as known in the art, for example, using the IgBLAST program at the NCBI site, as described in Ye et al. Nucleic Acids Res. 41: W34-W40 (2013), incorporated herein in its entirety by reference for humanization methods. IgBLAST takes a mouse VH and/or VL sequence and compares it to a library of known human germline sequences. As shown herein, for the humanized sequences generated herein, the databases used were IMGT human VH genes (F + ORF, 273 germline sequences) and IMGT human VL kappa genes (F + ORF, 74 germline sequences) . Exemplary five CHA sequences were selected: CHA.9.536, CHA9.560, CHA.9.546, CHA.9.547, and CHA.9.541 (see FIG. 53). For this humanization variant, human germline IGHV1-46 (allele 1) was chosen for all 5 as the acceptor sequence and binding region of the human IGHJ4 heavy chain (allele 1) (J-segment of the gene). For three of the four (CHA.7.518, CHA.7.530, CHA.7.538_1 and CHA.7.538_2) human germline IGKV1-39 (allele 1) was chosen as the acceptor sequence, and IGKJ2 (allele 1) (J-segment gene) human light chain was selected. The J-segment of the gene was selected from human fusion region sequences collected in the IMGT® international information system ImMunoGeneTics as www.imgt.org. CDRs were determined according to the AbM definition (see www.bioinfo.org.uk/abs/).

In some embodiments, the anti-TIGIT antibodies of the present invention include anti-TIGIT antibodies in which the V _H and V L sequences of different anti-TIGIT antibodies can be mixed and matched to create other anti-TIGIT antibodies. TIGIT binding of such mixed and matched antibodies can be tested using the binding assays described above, eg ELISA or Biacore assays). In some embodiments, when the VH and VL strands are mixed and matched, the V _H sequence from a particular V _H /V _L pairing is replaced with a structurally similar V _H sequence. Similarly, in some embodiments, the V _L sequence from a particular V _H /V _L pairing is replaced with a structurally similar V _L sequence. For example, the V _H and V _L sequences of homologous antibodies are particularly amenable to mixing and matching.

Accordingly, the anti-TIGIT antibodies of the invention comprise CDR amino acid sequences selected from the group consisting of (a) the sequences listed herein; (b) sequences that differ from the amino acid sequences of the CDRs indicated in (a) by 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more amino acid substitutions; (c) amino acid sequences having 90% or more, 95% or more, 98% or more, or 99% or more sequence identity with the sequences specified in (a) or (b); (d) a polypeptide having an amino acid sequence that is encoded by a polynucleotide having a nucleic acid sequence encoding the amino acids listed herein. In particular, the CPA.9.086 antibody may have sequences selected from (a), (b), (c) or (d).

In addition, the definition of anti-TIGIT antibodies includes antibodies that have an identity with the anti-TIGIT antibodies listed herein. That is, in some embodiments, an anti-TIGIT antibody of the invention comprises heavy and light chain variable regions comprising amino acid sequences that are identical to all or part of the anti-TIGIT amino acid sequences of the preferred anti-TIGIT antibodies, respectively, wherein the antibodies retain the desired functional properties. parent anti-TIGIT antibodies. The percent identity between two sequences is a function of the number of identical positions that the given sequences have in common (i.e. % homology = # of identical positions / total number of positions x 100), taking into account the number of gaps and the length of each gap that must be entered to optimally align the two sequences. Comparison of sequences and definitions

- 36 040773 Determining percent identity between two sequences can be performed using a mathematical algorithm as described in the non-limiting examples below.

The percent identity between two amino acid sequences can be determined using the algorithm of E. Meyers and W. Miller (Comput. Appl. Biosci., 4: 11-17 (1988)), which was included in the ALIGN program (version 2.0), using the table weights of PAM120 residue substitutions, a gap extension penalty of 12, and a gap insertion penalty of 4. In addition, percent identity between two amino acid sequences can be determined using Needleman and Wunsch (J. Mol. Biol. 48:444- 453 (1970)), which was included in the GAP program in the GCG software package (commercially available), using either the Blossum 62 matrix or the PAM250 matrix, and a gap penalty of 16, 14, 12, 10, 8, 6 or 4 and a gap extension penalty of 1, 2, 3, 4, 5, or 6.

Additionally or alternatively, the protein sequences of the present invention may additionally be used as a query sequence to perform open database searches, for example, to identify related sequences. Such searches can be performed using the XBLAST program (version 2.0) of Altschul, et al. (1990) J Mol. Biol. 215:403-10. BLAST protein searches can be performed using the XBLAST program, score=50, wordlength=3, to obtain amino acid sequences homologous to antibody molecules, in accordance with at least some embodiments of the invention. To obtain gapped alignments for comparison purposes, Gapped BLAST can be used as described in Altschul et al., (1997) Nucleic Acids Res. 25(17):3389-3402. When using the BLAST and Gapped BLAST programs, the default settings for the respective programs (eg XBLAST and NBLAST) may be used.

In general, the percent identity for comparison between anti-TIGIT antibodies is at least 75%, at least 80%, at least 90%, with at least 95%, 96%, 97%, 98%, or 99% being preferred. Percent identity may be along the entire amino acid sequence, such as the entire heavy or light chain, or along part of the chains. For example, within the definition of anti-TIGIT antibodies of the invention are those that share identity along the entire variable region (eg, where the identity is 95 or 98% along the variable regions) or along the entire constant region or along the Fc domain alone. In particular, the invention provides antibodies to TIGIT that have at least 75%, at least 80%, at least 90%, with a preferred identity of at least about 95%, 96%, 97%, 98%, or 99% identity with the CPA antibody. 9.086.

In addition, sequences are also included that may have identical CDRs but changes in the framework portions of the variable domain (or the entire heavy or light chain). For example, anti-TIGIT antibodies include antibodies with CDRs identical to those depicted in FIG. 53, but which may have lower identity along the variable region, such as 95% or 98% identity. In particular, the invention provides antibodies to TIGIT that have identical CDRs to CPA.9.086, but with framework regions that are 95% or 98% identical to CPA.9.086.

A. Anti-TIGIT antibodies that compete for binding

The present invention provides not only the listed antibodies, but also additional antibodies that compete with the listed antibodies (the CPA numbers listed herein that specifically bind to TIGIT) for specific binding to the TIGIT molecule. As shown in Example 16, the TIGIT antibodies of the invention are sorted into different epitope groups. Among the 44 anti-TIGIT antibodies in the epitope-specific sorting study, there are four communities, each with associated pairwise blocking patterns, which are divided into 12 general discrete groups described herein and depicted in FIG. 67 and 68. Twelve discrete groups are represented in this document;

1) VM9-H4, SHA.9.525, CPA.9.081-H4, SHA.9.538, SHA.9.553, CPA.9.069-H4,

CHA.9.543, CHA.9.556, CPA.9.077-H4 and CHA.9.561; 2) SHA.9.560 and SHA.9.528; 3)

CHA.9.552, CHA.9.521, CHA.9.541, CHA.9.529, CHA.9.519, CHA.9.527 and CHA.9.549;4)

CPA.9.057-H4 and CHA.9.554; 5) CHA.9.546, CPA.9.012-H4, CHA.9.547, CPA.9.013-H4,

CPA.9.018-H4, MBSA43-M1, Sino PVR-Fc (ligand), CHA.9.555, PVR-Fc M2A (ligand), BM29-H4, CPA.9.027-H4, CPA.9.049-H4 and CPA.9.053- H4; 6) CPA.9.064-H4; 7) VM26-H4; 8) CPA.9.059-H4; 9) CHA.9.535 and CPA.9.009-H4; 10) CHA.9.536, CHA.9.522 and CPA.9.015-H4;

11) CPA.9.011-H4 and BM8-H4 and 12) CPA.9.071-H4.

Thus, the invention provides anti-TIGIT antibodies that compete for binding to antibodies that are in discrete epitope groups 1 to 12. In a specific embodiment, the invention provides anti-TIGIT antibodies that compete for binding to CPA.9.086 and at least least 95, 96, 97, 98 or 99% identical to CPA.9.086.

Additional antibodies that compete with the listed antibodies are made as is known in the art and are generally described below. Competitive binding studies

- 37 040773 can be carried out, as is known in the art, usually using SPR/Biacore® binding assays, as well as using ELISA and cell assays.

VIII. Antibodies to PVRIG

The present invention provides anti-PVRIG antibodies (for convenience, anti-PVRIG antibodies and anti-PVRIG antibodies are used interchangeably). The anti-PVRIG antibodies of the invention bind specifically to human PVRIG and preferably to human PVRIG ECV.

Specific binding of PVRIG or a PVRIG epitope can be demonstrated, for example, by an antibody having a KD of at least about 10 ^-4 M, at least about 10 ^-5 M, at least about 10 ^-6 M, at least about 10 ^-7 M, at least about 10 ^-8 M, at least about 10 ^-9 M, at least about 10 ^-10 M, at least about 10 ^-11 M, at least about 10 ^-12 M or more, where KD refers to the rate of dissociation of a particular antibody-antigen interaction. Typically, an antibody that specifically binds an antigen will have a KD that is 20, 50, 100, 500, 1000, 5000, 10,000 or more times the control molecule relative to the PVRIG antigen or epitope.

However, as shown in the examples of WO2016/134333, for optimal binding to PVRIG expressed on the surface of NK and T cells, antibodies preferably have a KD of less than 50 nM and most preferably less than 1 nM, using less than 0.1 nM and less than 1 and 0.1 pM in the methods of the invention.

Specific binding of a particular antigen or epitope may also be shown, e.g., an antibody having KA or Ka for a PVRIG antigen or epitope at least 20, 50, 100, 500, 1000, 5000, 10,000 or more times for the epitope relative to a control, where KA or Ka refers to the rate of association of a particular antibody-antigen interaction.

In some embodiments, the anti-PVRIG antibodies of the invention bind to human PVRIG with a KD of 100 nM or less, 50 nM or less, 10 nM or less, or 1 nM or less (i.e., higher binding affinity), or 1 pM, or however, KD is determined by known methods, eg surface plasmon resonance (SPR, eg Biacore assays), ELISA (enzyme-linked immunosorbent assay), KINEXA (kinetic exclusion assay), and the most common SPR at 25 or 37°C.

It is important to note that the binding affinity for an anti-PVRIG antibody unexpectedly correlates with potency. Cumulative analysis of screening data demonstrates that the affinity of the anti-PVRIG antibodies of the invention is highly correlated with their ability to bind to primary human T cells. More specifically, the antibodies that produced the highest maximum signal on T cells were antibodies with affinity in the picomolar range. Antibodies that had an affinity in the low nanomolecular range and above produced relatively weak maximal signals on T cells. Thus, the data demonstrate that the usefulness of anti-PVRIG antibodies for T cell-based immunotherapy can be determined, in part, based on their affinity. Reference is made to the antibody sequences from WO 2016/134333 incorporated herein by reference and in particular to the anti-PVRIG antigen-binding domains shown in FIG. 38 (which depicts sequences that bind PVRIG and block the interaction of PVRIG and PVRL2), FIG. 39 (which depicts sequences that bind PVRIG and do not block the interaction of PVRIG and PVRL2), FIG. 40 (which depicts CDRs and data from these antibodies) and FIG. 41 which shows CDRs from hybridomas that bind and block). That is, the figures and conventions as well as specific sequences and SEQ ID NO: from all CPA.7 and CHA.7 antibodies (including CDRs, VH and VL and full length sequences) of WO 02016/134333 are expressly included in this document.

The drawing illustrates the ability of two anti-PVRIG antibodies with different affinities to bind primary CD8 T cells. As depicted in the figure, CHA.7.518 has about 8 times higher affinity than CPA.7.021 (sequence in WO 2016/13433) as measured by binding to HEK cells modified to overexpress PVRIG (HEK hPVRIG). Consistent with this, CHA.7.518 has an approximately 13-fold higher affinity than CPA.7.021 as measured by binding to Jurkat cells. The higher affinity of CHA.7.518 corresponded to a higher maximum binding signal from HEK hPVRIG cells, but not from Jurkat cells.

In contrast, CHA.7.518 consistently produced a higher maximum binding signal to primary CD8 T cells compared to CPA.7.021. This is illustrated in a binding titration experiment in which various concentrations of isotype or anti-PVRIG antibodies were added to primary CD8 T cells and the maximum binding signal obtained was measured. In the two donors depicted, CHA.7.518 consistently produced a higher maximum signal (geometric mean fluorescence intensity, gSIF) than CPA.7.021, depending on the titration. gMFIr = geometric mean fluorescence intensity of the antibody of interest/geometric fluorescence intensity of the control antibody. The specified gMFIr measures the signal of the antibody of interest relative to the isotype antibody at a fixed

- 38 040773 concentrations of both.

Accordingly, anti-PVRIG antibodies of the invention have a binding affinity (as measured using the methods described herein) in the picomolar range, for example, from 0.1 to 9 pM, with about 0.2 to about 2 being preferred, and about 0.2 to about 0.5 for practical use.

With regard to antibodies to TIGIT, antibodies to PVRIG are similarly labeled as follows. Antibodies have code numbers such as CHA.7.518.1. This is a combination of heavy and light chain variable regions as shown in FIG. , for example, with the understanding that these antibodies contain two heavy chains and two light chains. SRA. 7.518.1.VH refers to the heavy chain variable portion of CPA.7.518.1 while CPA.7.518.1.VL is the light chain variable portion. CPA.7.518.1.vhCDR1, CPA.7.518.1.vhCDR2, CPA.7.518.1.vhCDR3, CPA.

7.518.1. VlCDR1, CPA. 7.518.1.VLCDR2, and CPA. 7.518.1.VLCDR3 refers to CDRs that are specified by the CPA. 7.518.1.HC refers to the entire heavy chain (eg, variable and constant domain) of that molecule and the CPA. 7.518.1.LC refers to the entire light chain (eg, variable and constant domain) of the same molecule. In general, a human kappa light chain is used for the constant domain of each phage antibody (or humanized hybridoma) herein, although in some embodiments, a lambda light chain constant domain is used. SRA. 7.518.1.H1 refers to a full-length antibody containing the variable domains of the heavy and light chains, including the human (Human) IgG1 constant domain (hence H1; the sequences of IgG1, IgG2, IgG3 and IgG4 are shown in the drawing). Accordingly, SRA. 7.518.1.H2 will be the CPA variable domains. 7.518.1 associated with human (Human) IgG2. SRA. 7.518.1.H3 will be the CPA variable domains. 7.518.1 associated with human (Human) IgG3, and CPA. 7.518.1.H4 will be the CPA variable domains. 7.518.1 associated with human (Human) IgG4. Note that in some cases human IgGs may have additional mutations, which are described below and this will be annotated. For example, in many embodiments, there may be an S241P mutation in human IgG4 and this may be labeled CPA.7.518.1.H4(S241P), for example. The sequence of human IgG4 with this hinged S241P variant is depicted in the figure. Other possible variants are IgG1 (N297A) (or other variants that lack glycosylation at this site and hence many of the effector functions associated with FcyRIIIa binding) and IgG1 (D265A), which reduces binding to FcyR receptors.

The invention further provides heavy and light chain variable domains as well as full length heavy and light chains.

In some embodiments, the invention provides scFvs that bind to a PVRIG containing a heavy chain variable domain and a light chain variable domain linked by an scFv linker as described above. The VL and VH domains can be in any orientation, such as N- to C-terminus of VH-linker-VL or VL-linker-VH. They are named after their component parts; for example, scFv-CHA.7.518.1VH-linker-VL or scFv-CPA. 7.518.1.VL-linker-VH. Thus, scFv-CPA. 7.518.1 can be in any orientation.

IX. Nucleic acids encoding antibodies

Also provided are nucleic acid compositions encoding antibodies of the invention, as well as expression vectors containing nucleic acids and host cells transformed using nucleic acid compositions and/or expression vectors. As will be understood by those skilled in the art, the protein sequences described herein may be encoded by any number of possible nucleic acid sequences due to the degeneration of the genetic code.

The nucleic acid compositions that encode antibodies will depend on the format of the antibody. For standard tetrameric antibodies containing two heavy chains and two light chains that are encoded by two different nucleic acids, one of which encodes the heavy chain and the other encodes the light chain. They can be placed in a single expression vector or two expression vectors, as is known in the art, transformed into host cells where they are expressed to form the antibodies of the invention. In some embodiments, for example when scFv constructs are used, a single nucleic acid encoding a heavy chain-linker-light chain variable region is typically used, which can be inserted into an expression vector for transformation into host cells. Nucleic acids can be introduced into expression vectors that contain appropriate transcriptional and translational control sequences, including, but not limited to, signal and secretory sequences, regulatory sequences, promoters, origins of replication, selection genes, etc.

Preferred mammalian host cells for expressing recombinant antibodies in accordance with at least some embodiments of the invention include Chinese Hamster Ovary (CHO cells), PER.C6, HEK293, and others as is known in the art.

Nucleic acids may be present in whole cells, in cell lysate, or in partial

- 39 040773 purified or essentially pure form. The nucleic acid is isolated or brought to substantially pure form when purified from other cellular components or other contaminants, e.g., other cellular nucleic acids or proteins, by standard methods, including alkali/SDS treatment, CsCl separation, column chromatography, agarose gel electrophoresis, and others. well known in the art.

To create an scFv gene, DNA fragments that encode VH and VL are operably linked to another fragment encoding a flexible linker, such as encoding the amino acid sequence (Gly4-Ser) 3 and others discussed herein, so that the V _H and V _L sequences can be expressed as a continuous single chain protein, with the V _L and V _H regions connected by a flexible linker.

X. Compositions

Therapeutic compositions used in the practice of the above methods (and in particular CHA.7.518.1.H4 (S241P) and CPA.9.086) can be formulated into pharmaceutical compositions containing a carrier suitable for the desired mode of delivery. Suitable carriers include any material which, in combination with the therapeutic composition, retains the antitumor function of the therapeutic composition and does not generally react with the patient's immune system. Examples include, but are not limited to, any of a number of standard pharmaceutical carriers such as sterile sodium phosphate buffer solutions, bacteriostatic water, and the like (see generally Remington's Pharmaceutical Sciences ^16th Edition, A. Osal, Ed., 1980 ). Acceptable carriers, excipients, or stabilizers are non-toxic to recipients at the doses and concentrations used, and may include buffers.

In a preferred embodiment, the pharmaceutical composition that contains the antibodies of the invention may be in a water-soluble form, such as present in the form of pharmaceutically acceptable salts, which is intended to include an acid addition salt and a base. A pharmaceutically acceptable acid addition salt refers to those salts which retain the biological effectiveness of free bases and which are not biologically or otherwise undesirable by inorganic acids and the like. Pharmaceutically acceptable base addition salts include those derived from inorganic bases and the like.

The introduction of a pharmaceutical composition containing antibodies according to this invention, preferably in the form of a sterile aqueous solution, can be carried out in various ways, including, but not limited to, subcutaneously and intravenously.

Dose amounts and frequencies of administration in a preferred embodiment are selected for therapeutic or prophylactic efficacy. As is known in the art, adjustments for protein degradation, systemic and local delivery, and the rate of new protease synthesis, as well as age, body weight, general health, sex, diet, time of administration, drug interactions, and disease severity can be necessary and will be determined using standard experiments by those skilled in the art.

A therapeutically effective dose of the Fc variant of the invention may be administered to treat a patient. A therapeutically effective dose is herein understood to mean a dose that produces the effects for which it is administered. The exact dose will depend on the goal of treatment and will be determined by one of skill in the art using known techniques.

XI. Ways to Use Antibodies

The antibodies of the invention, including both PVRIG and TIGIT antibodies, can be used in a variety of diagnostic and therapeutic applications. In some cases, the decision on which antibody to administer to a patient is made using an assessment of expression levels (gene expression levels or protein expression levels, the latter being preferred) of biopsies of tumor specimens to determine whether the specimen overexpresses either TIGIT or PVRIG, or and both to determine which therapeutic antibodies to administer.

A. Diagnostic application

Accordingly, the antibodies of the invention also find use in in vitro or in vivo diagnostics, including imaging of tumors that overexpress either PVRIG or TIGIT, respectively. It should be noted, however, that as discussed herein, both TIGIT and PVRIG, as immuno-oncological target proteins, are not necessarily overexpressed on cancer cells, but rather in cancer immune infiltrates. Thus, the mechanism of action, such as the activation of immune cells such as T cells and NK cells, determines the diagnosis of cancer. Accordingly, these antibodies can be used to diagnose cancer. Diagnosis using antibodies to PVRIG is also described in WO 2016/134333, [0434-0459], incorporated herein by reference.

As a rule, diagnostics can be performed in several ways. In one embodiment, tissue from a patient, such as a biopsy sample, is contacted with a typically labeled anti-TIGIT antibody such that the antibody binds to endogenous TIGIT. Signal level ref.

- 40 040773 are compared with the level of normal non-cancerous tissue either in the same patient or with a reference sample to determine the presence or absence of cancer. The biopsy sample can be obtained from a solid tumor, a blood sample (for lymphomas and leukemias such as ALL (acute lymphoblastic leukemia), T-cell lymphoma, etc.).

In general, in this embodiment, the anti-TIGIT is labeled with, for example, a fluorophore or other optical label that is detected using a fluorometer or other optical detection system, as is well known in the art. In an alternative embodiment, a secondary labeled antibody is contacted with a sample, e.g. using an anti-human IgG antibody from another mammal (mouse, rat, rabbit, goat, etc.), to perform a sandwich assay, as is known in the art. technology. Alternatively, the anti-TIGIT mAb can be directly labeled (ie, with biotin) and detection can be performed by a secondary Ab directed to the labeled agent in the art.

Once TIGIT overexpression is determined, treatment may be continued with the administration of an anti-TIGIT antibody of the invention as described herein.

In other embodiments, in vivo diagnostics are performed. Typically, in this embodiment, an anti-TIGIT antibody (including antibody fragments) is administered to a patient and imaging is performed. In this embodiment, for example, the antibody is typically labeled with an optical or MRI label such as gadolinium chelate, mAb radiolabeling (including fragments).

In some embodiments, the antibodies described herein are used for both diagnosis and treatment, or only for diagnosis. When anti-TIGIT antibodies are used for both diagnosis and treatment, some embodiments rely on two different anti-TIGIT antibodies for two different epitopes so that the diagnostic antibody does not compete for binding with the therapeutic antibody, although in some cases the same the same antibody can be used for both cases. For example, this can be done using antibodies that are in different groups, for example, that bind to different epitopes on TIGIT, for example, as described in this document. Thus, compositions comprising a diagnostic antibody and a therapeutic antibody are included in the invention, and in some embodiments, the diagnostic antibody is labeled as described herein. In addition, the composition of therapeutic and diagnostic antibodies can also be administered together with other drugs, as described in this document.

Particularly useful antibodies for use in diagnostics include, but are not limited to, these antibodies listed, or antibodies that use CDRs with variant sequences, or those that compete for binding to any of the antibodies of Figure 1.

In many embodiments, the diagnostic antibody is labelled. By labeled herein is meant that the antibodies disclosed herein contain one or more elements, isotopes, or chemical compounds attached to allow detection in an assay or diagnostic procedure. In general, labels fall into several classes: a) immune labels, which can be an epitope included as a fusion partner that is recognized by an antibody; b) isotopic markers, which may be radioactive or heavy isotopes; c) small molecular weight labels, which may include fluorescent and colorimetric dyes, or molecules such as biotin, which allow other labeling methods, and d) labels, such as particles (including vials for ultrasonic labeling) or paramagnetic labels, which enable imaging in the body . These labels can be incorporated into antibodies at any position and can be introduced in vitro or in vivo during protein expression, as is known in the art.

Diagnosis can be made either in vivo or by administering a diagnostic antibody that allows whole body imaging as described below, or in vitro on specimens removed from the patient. A sample in this context includes any number of things including, but not limited to, bodily fluids (including, but not limited to, blood, urine, serum, lymph, saliva, anal and vaginal secretions, sweat, and semen), and tissue samples. , such as the result of a biopsy of the relevant tissues.

In addition, as shown below and in the examples and figures, information about the protein expression levels of either PVRIG or TIGIT, or both, or PVRIG and PD-1, or TIGIT and PD-1, can be used to determine which antibodies should be administered to the patient.

B. Cancer treatment

The antibodies of the invention are particularly useful in the treatment of cancer. In general, the antibodies of the invention are immunomodulatory, in that instead of directly attacking cancer cells, the antibodies of the invention stimulate the immune system, typically by inhibiting the action of a checkpoint receptor (eg, PVRIG or TIGIT). Thus, unlike tumor-specific therapies that aim to inhibit molecular pathways that are critical for tumor growth and development and/or tumor cell depletion, cancer immunotherapy aims to stimulate the patient's own immune system to eliminate cancer cells, ensuring the destruction of resistant tumors. . Can be used in cancer therapy

- 41 040773 personal approaches, including therapeutic cancer vaccines to induce tumor-specific T cell responses and immunostimulatory antibodies (ie inhibitory receptor antagonists = immune checkpoints) to remove immunosuppressive pathways.

Clinical responses with targeted therapy or conventional cancer therapy are generally temporary as cancer cells develop resistance and tumor recurrence occurs. However, the clinical use of cancer immunotherapy in the past few years has shown that this type of therapy can have long-lasting clinical responses, demonstrating a strong impact on long-term survival. However, although responses are long-term, only a small number of patients respond (as opposed to conventional or targeted therapy where a large number of patients respond but the responses are temporary).

By the time a tumor is detected clinically, it has already evaded the immune defense system by acquiring immunoresistant and immunosuppressive properties and creating an immunosuppressive tumor microenvironment through various mechanisms and various immune cells.

Accordingly, the antibodies of the invention are useful in the treatment of cancer. Due to the nature of the immuno-oncological mechanism of action, the checkpoint receptor (TIGIT or PVRIG) need not be overexpressed or correlated with a particular type of cancer; that is, the goal is for the antibodies to remove the suppression of T cell and NK cell activation so that the immune system responds to cancer.

Cancer as used herein refers generally to any neoplastic disease (invasive or metastatic) characterized by abnormal and uncontrolled cell division causing malignant growth or tumor (eg, unregulated cell growth). The term cancer or malignant as used herein should be understood to encompass any neoplastic disease (invasive, non-invasive, or metastatic) that is characterized by abnormal and uncontrolled cell division causing malignant growth or tumor, non-limiting examples of which are described herein. This includes any physiological condition in mammals that is typically characterized by unregulated cell growth. Examples of cancer are given in the working examples and are also described in the description.

Non-limiting examples of cancers that can be treated using the antibodies of the invention include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia. More specific examples of such cancers include squamous cell carcinoma, lung cancer (including small cell lung cancer, non-small cell lung cancer, lung adenocarcinoma, and squamous cell lung cancer), abdominal cancer, hepatocellular gastric cancer, or gastric cancer (including gastrointestinal cancer). tract), pancreatic cancer, glioblastoma, cervical cancer, ovarian cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, colorectal cancer, endometrial cancer or uterus, salivary gland carcinoma, kidney or renal cancer, liver cancer, prostate cancer, vulvar cancer, thyroid cancer, liver carcinoma, and various types of head and neck cancer, and B-cell lymphoma (including slow-growing/follicular non-Hodgkin's lymphoma). lymphoma (NHL), small cell lymphocytic (ML) NHL, moderately aggressive/follicular NHL, moderately aggressive diffuse NHL, fast growing immunoblastic NHL, fast growing lymphoblastic NHL, fast growing small cell NHL, generalized NHL; mantle cell lymphoma; AIDS-associated lymphoma; and macroglobulinemia of Waldenström); chronic lymphocytic leukemia (CLL); acute lymphoblastic leukemia (ALL); hairy cell leukemia; chronic myeloid leukemia; multiple myeloma; and post-transplant lymphoproliferative disorder (PTLD).

As shown in the examples of WO 2016/134333, PVRIG is overexpressed and/or correlated with tumor lymphocyte infiltration (as evidenced by correlation with CD3, CD4, CD8 and PD-1 expression) in a number of different tumors of various origins and thus useful in the treatment of any cancer. including but not limited to prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, gastric cancer, cervical cancer, head and neck cancer, thyroid cancer , testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (non-Hodgkin's lymphoma (NHL) and Hodgkin's lymphoma (HD)), acute myeloid leukemia (AML) ), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular germ cell tumors, mesothelioma, and esophageal cancer.

In particular, CHA.7.518.SH4 (S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, stomach cancer, cervical cancer, cancer head and neck, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia (AML) , T-cell acute lymphoblastic leukemia (T-ALL), diffuse large cell B

- 42 040773 cellular lymphoma, testicular germ cell tumors, mesothelioma, bladder cancer and esophageal cancer.

In particular, CHA.7.538.1.2.H4(S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, gastric cancer, cervical cancer. , head and neck cancer, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia ( AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular tumors derived from terminal cells, mesothelioma, bladder cancer and esophageal cancer.

In particular, CPA.9.086H4(S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, stomach cancer, cervical cancer, head cancer. and neck, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia (AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular tumors derived from terminal cells, mesothelioma, bladder cancer and esophageal cancer.

In particular, CPA.9.083H4(S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, stomach cancer, cervical cancer, head cancer. and neck, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia (AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular germ cell tumors, mesothelioma, bladder cancer, and esophageal cancer.

In particular, CHA.9.547.7.H4(S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, gastric cancer, cervical cancer. , head and neck cancer, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia ( AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular germ cell tumors, mesothelioma, bladder cancer, and esophageal cancer.

In particular, CHA.9.547.13.H4(S241P) is used in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, gastric cancer, cervical cancer. , head and neck cancer, thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia ( AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular germ cell tumors, mesothelioma, bladder cancer, and esophageal cancer.

C. Anti-TIGIT Antibody Monotherapy

Anti-TIGIT antibodies of the invention are particularly useful in the treatment of cancer as monotherapy. Due to the nature of the immuno-oncological mechanism of action, TIGIT need not be overexpressed or correlated with a particular type of cancer; that is, the goal is for the anti-TIGIT antibodies to remove the suppression of T-cell and NK-cell activation so that the immune system responds to oncological diseases.

While any anti-TIGIT antibody of FIG. 53 finds use in the treatment of cancer (including T cell activation as noted below), CPA.9.086.H4 (S241P), CPA.9.083.H4 (S241P), CHA.9.547.7.H4 (S241P), and CHA. 9.547.13.H4 (S241P) finds particular use in some embodiments. Fig.

D. Anti-PVRIG Antibody Monotherapy

Anti-PVRIG antibodies of the invention are particularly useful in the treatment of cancer as monotherapy. Due to the nature of the immuno-oncological mechanism of action, TIGIT need not be overexpressed or correlated with a particular type of cancer; that is, the goal is for anti-TIGIT antibodies to de-suppress T cells and NK cells so that the immune system responds to cancer.

In particular, CHA.7.518.1H4 (S241P) finds use as a monotherapy.

Similarly, CHA.7.538.1.2.H4 (S241P) finds use as monotherapy in the treatment of prostate cancer, liver cancer (HCC), colorectal cancer, ovarian cancer, endometrial cancer, breast cancer, pancreatic cancer, gastric cancer, cancer cervical cancer, head and neck cancer,

- 43 040773 thyroid cancer, testicular cancer, urothelial cancer, lung cancer, melanoma, non-melanoma skin cancer (squamous cell carcinoma and basal cell carcinoma), glioma, kidney cancer (RCC), lymphoma (NHL or CL), acute myeloid leukemia (AML), T-cell acute lymphoblastic leukemia (T-ALL), diffuse large B-cell lymphoma, testicular germ cell tumors, mesothelioma, bladder cancer, and esophageal cancer.

E. Combination Therapies

As is known in the art, combination therapies containing a therapeutic antibody targeted to an immunotherapeutic target and an additional disease-specific therapeutic agent show great promise. For example, in the field of immunotherapy, there are a number of promising combination therapies using a chemotherapeutic agent (either a small molecule drug or an anti-tumor antibody) or an immuno-oncological antibody.

The terms in combination with and co-administration are not limited to administering said prophylactic or therapeutic agents at exactly the same time. Instead, it is intended that the antibody and other agent or agents be administered in sequence and over time so that they can act together to provide a benefit that is increased over treatment with an antibody of the invention or another agent or agent alone. It is preferred that the antibody and the other agent or agents act additively, and it is particularly preferred that they act synergistically. Such molecules are suitably present in combination in amounts that are effective for the intended purpose. The skilled physician can determine, empirically or considering the pharmacokinetics and modes of action of the agents, the appropriate dose or doses of each therapeutic agent, as well as the appropriate timing and routes of administration.

Accordingly, the antibodies of this invention may be administered simultaneously with one or more other treatment regimens or agents. Additional treatment regimens or agents may be used to improve the efficacy or safety of the antibody. In addition, additional treatment regimens or agents may be used to treat the same disease or comorbidity, and not to change the action of the antibody. For example, an antibody of the invention may be administered to a patient in conjunction with chemotherapy, radiation therapy, or both chemotherapy and radiation therapy.

1. Antibodies to TIGIT with small molecular weight chemotherapeutic substances

The anti-TIGIT antibodies of this invention may be administered in combination with one or more other prophylactic or therapeutic agents, including, but not limited to, cytotoxic agents, chemotherapeutic agents, cytokines, growth inhibitory agents, antihormonal agents, kinase inhibitors, antiangiogenic agents, cardioprotective agents, immunostimulatory agents, immunosuppressants, agents that promote hematological cell proliferation, angiogenesis inhibitors, protein tyrosine kinase (PTK) inhibitors, or other therapeutic agents.

In this context, a chemotherapeutic agent is a chemical compound useful in the treatment of cancer. Examples of chemotherapeutic agents include alkylating agents such as thiotepa or cyclophosphamide; alkylsulfonates such as busulfan, improsulfan and piposulfan; aziridines such as benzodopa, carbokwon, meturedopa and uredopa; ethyleneimines and methylamelamines, including altretamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramide, and trimethylomelamin; acetogenins (especially bullatacin and bullatacinone); delta-9tetrahydrocannabinol (dronabinol, MARINOL'); beta-lapachone; lapachol; colchicines; betulinic acid; camptothecin (including the synthetic analogue of topotecan (HYCAMTN®), CPT-11 (irinotecan, CAMPTOSAR®), acetylcamptothecin, scopolectin and 9-aminocamptothecin); bryostatin; callistatin; CC-1065 (including its synthetic analogues adozelesin, carzelesin and bizelesin); podophyllotoxin; podophyllic acid; teniposide; cryptophycins (in particular, cryptophycin 1 and cryptophycin 8); dolastatin; duocarmycin (including synthetic analogues of KW-2189 and CB1-TM1); eleutherobin; pancratistatin; sarcodictin; spongistatin; nitrogen mustards such as chlorambucil, chlornaphasine, chlorophosphamide, estramustine, ifosfamide mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembichin, phenesterin, prednimustine, trofosfamide, uracil mustard; nitrosoureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine and ranimnustine; antibiotics such as enediine antibiotics (eg calicheamicin, especially calicheamicin gamma 11 and calicheamicin omega 11 (see, for example, Agnew, Chem Intl. Ed. Engl, 33: 183-186 (1994)); dynemycin, including dynemycin A; esperamycin; as well as the neocarcinostatin chromophore and related chromoprotein endien antibiotic chromophores), aclacinomysins, actinomycin, autramycin, azaserine, bleomycins, cactinomycin, carabicin, carminomycin, carcinophilin, chromomycins, dactinomycin, daunorubicin, detorubicin, 6-oxo-R norleucine, doxorubicin (including morpholinodoxorubicin, cyanomorpholinodoxorubicin, 2-pyrrolinodoxorubicin, and deoxydoxorubicin), epirubicin,

- 44 040773 esorubicin, idarubicin, marcellomycin, mitomycins such as mitomycin C, mycophenolic acid, nogalamycin, olivomycins, peplomycin, potfiromycin, puromycin, quelamycin, rhodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-fluorouracil (5-FU); folic acid analogs such as denopterin, methotrexate, pteropterin, trimetrexate; purine analogs such as fludarabine, 6-mercaptopurine, thiamiprin, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxyfluridine, enocytabine, floxuridine; androgens such as calusterone, dromostanolone propionate, epitiostanol, mepitiostane, testolactone, adrenal depressants such as aminoglutethimide, mitotane, trilostane; folic acid compensators such as folinic acid; aceglatone; aldophosphamide glycoside; aminolevulinic acid; eniluracil; amsacrine; bestrabucil; bisantrene; edatraksat; defofamine; demecolicin; diaziquon; elfornitine; elliptium acetate; epothilone; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidanin; maytansinoids such as maytansine and ansamitocins; mitoguazone; mitoxantrone; mopidanmol; nitraerin; pentostatin; phenamet; pyrarubicin; losoxantrone; 2-ethylhydrazide; procarbazine; polysaccharide complex PSK.RTM (JHS Natural Products, Eugene, OReg); razoxane; rhizoxin; sisofiran; spirogermanium; tenuazonic acid; triaziquone; 2,2',2-trichlorotriethylamine; trichothecenes (especially T-2 toxin, verracurin A, roridin A and anguidin); urethane; vindesine (ELDISINE®, FILDESIN®); dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; hacytosine; arabinoside (Ara-C); thiotepa; taxoids such as paclitaxel (TAXOL®; Bristol-Myers Squibb Oncology, Princeton, N.J.), ABRAXANE®, paclitaxel nanoparticle albumin formulation (American Pharmaceutical Partners, Schaumberg, Ill.), and docetaxel (TAXOTERE®; Rhone- Poulenc Rorer, Antony, France)); chloranbucil; gemcitabine (GEMZARM®); 6-thioguanine; mercaptopurine; methotrexate; platinum analogues such as cisplatin and carboplatin; vinblastine (VELBAN®); platinum; etoposide (VP-16); ifosfamide; mitoxantrone; vincristine (ONCOVIN®); oxaliplatin; folinic acid; vinorelbine (NAVELBINE®); novantron; edatrexate; daunomycin; aminopterin; ibandronate; topoisomerase inhibitor RFS 2000; difluoromethylornithine (DMFO); retinoids such as retinoic acid; capecitabine (XELODA®); their pharmaceutically acceptable salts, acids or derivatives of any of the above, as well as combinations of two or more of the above, such as CHOP is an abbreviation for combination therapy with cyclophosphamide, doxorubicin, vincristine and prednisolone; CVP is an abbreviation for combination therapy of cyclophosphamide, vincristine, and prednisolone; and FOLFOX is an abbreviation for an oxaliplatin (ELOXATIN®) regimen in combination with 5-FU and leucovorin.

In at least some embodiments, anti-TIGIT immune molecules can be used in combination with any of the art-known cancer treatment standards (as can be found, for example, at http://www.cancer.gov/cancertopics).

Thus, in some cases, the anti-PVRIG antibodies described herein (in particular, including CHA.7.538.1.2.H4 (S241P) or CHA.7.518.1.H4 (S241P)) can be combined with chemotherapeutic agents. Similarly, the anti-TIGIT antibodies described herein (in particular, including CPA.9.086H4 (S241P), CPA9.083H4 (S241P) and CHA.9.547.13.H4 (S241P)) can be combined with chemotherapeutic agents.

In addition, the anti-PVRIG and anti-TIGIT antibodies of the invention may also be administered with other checkpoint inhibitors or activators.

2. Combination therapy with anti-TIGIT and anti-checkpoint antibodies

As shown herein, the TIGIT antibodies of the invention can be combined with one of several checkpoint antibodies. In some embodiments, a patient's tumor can be assessed for receptor expression and the results are then used to inform the clinician which antibodies to use: PVRIG and PD-1, TIGIT and PD-1, or TIGIT and PVRIG. These assays are described below.

a) Anti-TIGIT antibodies in combination with anti-PD-1 antibodies

In one embodiment, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-PD-1 antibodies.

In one embodiment, the biopsy is taken from a tumor in a cancer patient and dissociated as known in the art for FACS analysis. Cells are stained with labeled antibodies to (1) TIGIT (eg, using any of those described herein or others in the art, such as MBSA43); (2) PD-1 (eg, using those known in the art, including EH12.2H7, Keytruda®, Opdivo®, etc.); (3) PD-L1 (e.g. using those known in the art such as BM-1 described herein) and (4) PVR (e.g. using those known in the art such as SKII.4) ; and (5) an isotype control antibody. FACS is done and for each receptor the percentage of cells expressing the receptor relative to the control antibody is calculated. If the percentage of positive cells for TIGIT, PD-1, PD-1 and PVR is > 1% for all 4 receptors, then the patient is treated with antibodies to TIGIT and PD-1 as described in this document.

- 45 040773

Accordingly, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-PD-1 antibodies. There are two approved anti-PD-1 antibodies, pembrolizumab (Keytruda®) and nivolumab (Opdivo®) and many others in development, that can be used in combination with the anti-TIGIT antibodies of the invention.

Accordingly, the invention provides specific combinations: CPA.9.083.H4(S241P) (as depicted in Fig. 53B) with pembrolizumab; CPA.9.083.H4 (S241P) as shown in FIG. 53B with nivolumab; CPA.9.086.H4(S241P) as shown in FIG. 53A with pembrolizumab; CPA.9.086.H4(S241P), as shown in FIG. 53A with nivolumab; CHA.9.547.7H4(S241P) with pembrolizumab; CHA.9.547.7H4(S241P with nivolumab, CHA.9.547.13.H4(S241P) with pembrolizumab, and CHA.9.547.13.H4(S241P) with nivolumab (Link to sequence listing).

b) Anti-TIGIT antibodies in combination with anti-CTLA-4 antibodies

In another embodiment, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-CTLA-4 antibodies. There are two approved anti-CTLA-4 antibodies, ipilimumab (Yervoy®) and tremelimumab, as well as others in development, that can be used in combination with the anti-TIGIT antibodies of the invention.

Accordingly, the invention provides specific combinations: CPA.9.083.H4(S241P) with ipilimumab; CPA.9.083.H4(S241P) with tremelimumab; CPA.9.086.H4 (S241P) with ipilimumab; CPA.9.086.H4(S241P) with tremelimumab; CHA.9.547.7H4(S241P) with ipilimumab; CHA.9.547.7H4(S241P) with tremelimumab; CHA.9.547.13.H4(S241P) with ipilimumab and CHA.9.547.13.H4(S241P) with tremelimumab.

c) Anti-TIGIT antibodies in combination with anti-PD-L1 antibodies.

In another embodiment, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-PD-L1 antibodies. There are three approved anti-PD-L1 antibodies, atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®) and durvalumab (IMFINZI™), as well as other anti-PD-L1 antibodies in development that can be used in combination with anti-TIGIT antibodies. according to the invention.

Accordingly, the present invention provides specific combinations of: CPA.9.083.H4(S241P) with atezolizumab; CPA.9.083.H4(S241P) with avelumab; CPA.9.083.H4(S241P) with durvalumab; CPA.9.086.H4(S241P) with atezolizumab; CPA.9.086.H4(S241P) with avelumab; CPA.9.086.H4(S241P) with durvalumab; CHA9.547.7H4(S241P) with atezolizumab; CHA9.547.7H4(S241P) with avelumab; CHA9.547.7H4(S241P) with durvalumab; CHA9.547.13.H4(S241P) with atezolizumab; CHA9.547.13.H4(S241P) with avelumab; and CHA9.547.13.H4(S241P) with durvalumab.

d) Anti-TIGIT antibodies in combination with Ahtu-LAG-3 antibodies

In another embodiment, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-LAG-3 antibodies. Several anti-LAG-3 antibodies are in development, including BMS-986016 (see International Patent Application No. WO 2010/019570A2, incorporated herein by reference) GSK2831781 (see US Patent Application No. 2016/0017037A, incorporated herein by reference) and clones Merck 22D2, 11C9, 4A10 and/or 19E8 (see WO 2016/028672A1, incorporated herein by reference in its entirety) and GSK2831781, as well as others in development that can be used in combination with anti-TIGIT antibodies according to the invention.

Accordingly, the invention provides specific combinations of:

CPA.9.083.H4(S241P) with BMS-986016; CPA.9.083.H4(S241P) with GSK2831781;

CPA.9.086.H4(S241P) with BMS-986016; CPA.9.086.H4(S241P) with GSK2831781;

CHA.9.547.7H4(S241P) with BMS-986016; CHA.9.547.7H4(S241P) with GSK2831781;

CHA.9.547.13.H4(S241P) with BMS-986016 and CHA.9.547.13.H4(S241P) with GSK2831781.

Accordingly, this invention also provides specific combinations: CPA9.083.H4(S241P) with Merck clones 22D2, 11C9 and/or 4A10; CPA9.086.H4(S241P) with Merck clones 22D2, 11C9 and/or 4A10; CHA.9.547.7H4(S241P) with Merck clones 22D2, 11C9 and/or 4A10; CHA9.547.13.H4(S241P) with Merck clones 22D2, 11C9 and/or 4A10.

f) Anti-TIGIT antibodies in combination with Lnti-T1M-3 antibodies

In another embodiment, the invention provides combinations of anti-TIGIT antibodies of the invention and anti-TIM-3 antibodies. There is at least one anti-TIM-3 antibody in development, TSR-022, as well as others in development that can be used in combination with anti-TIGIT antibodies of the invention.

Accordingly, the invention provides specific combinations: CPA.9.083.H4(S241P) with TSR022; CPA.9.086.H4(S241P with TSR-0226, CHA.9.547.7H4(S241P) with TSR-022 and CHA.9.547.13.H4(S241P) with TSR-022.

f) Anti-TIGIT antibodies in combination with anti-BTLA antibodies

In another embodiment, the invention provides combinations of anti-TIGIT antibodies according to the invention and anti-BTLA antibodies, see WO 2011/014438, incorporated by reference in

- 46 040773 in its entirety, and in particular for the CDRs and full-length sequences of the anti-BTLA antibodies described therein. Accordingly, the invention provides specific combinations of: CPA.9.083.H4(S241P) with an anti-BTLA antibody; CPA.9.086.H4(S241P) with anti-BTLA antibody; CHA9.547.7H4(S241P) with anti-BTLA antibody; and CHA.9.547.13.H4(S241P with anti-BTLA antibody.

g) TIGIT antibodies with antitumor antibodies

In some embodiments, the anti-TIGIT antibodies of the invention are co-administered with antibodies that, unlike immuno-oncological/checkpoint inhibitors that typically act on the immune system to increase the patient's native immune response, are instead directed against a specific tumor target antigen (TTA ). There are a large number of anti-TTA antibodies approved or in development that can be combined with these anti-TIGIT antibodies. Currently approved antibodies include, but are not limited to, cetuximab, panitumumab, nimotuzumab (all for EGFR), rituximab (CD20), trastuzumab and pertuzumab (HER2), alemtuzumab (CD52), bevacizumab (VEGF), ofatumumab (CD20), denosumab (RANK ligand), brentuximab (CD30), daratumumab (CD38), ibritumumab (CD20), and ipilimumab (CTLA4). Specific targeted oncology antibodies in clinical trials that can be combined with anti-TIGIT antibodies include, but are not limited to, Ahtu-CTLA4 mAbs such as ipilimumab, tremelimumab; anti-PD-1 such as nivolumab BMS-936558/MDX-1106/ONO4538, AMP224, CT-011, MK-3475, α anti-PDL-1 antagonists such as BMS-936559/MDX-1105, MEDI4736, RG- 7446/MPDL3280A; anti-LAG-3 such as IMP-321), anti-TIM-3, anti-BTLA, anti-B7H4, anti-B7-H3, anti-VISTA; Agonistic antibodies targeting immunostimulatory proteins, including anti-CD40 mAbs such as CP-870,893, lucatumumab, dacetuzumab; anti-CD137 such as BMS-663513 urelumab (anti-4-1BB, see, for example, US Pat. Nos. 7,288,638 and 8,962,804, incorporated herein by reference in its entirety); PF-05082566 utomilumab (see, for example, US Patent Nos. 8821867; 8337850; and 9468678 and International Patent Application Publication No. WO 2012/032433, incorporated herein by reference in its entirety); anti-OX40 mAb, such as anti-OX40 (see, for example, /029879 or WO2010096418, incorporated herein by reference in its entirety); anti-GITR mAbs such as TRX518 (see, for example, US Pat. No. 7,812,135, incorporated herein by reference in its entirety); anti-CD27 mAb, such as varlilumab CDX-1127 (see, for example, WO 2016/145085 and US Patent Publication No. US 2011/0274685 and US 2012/0213771, incorporated herein by reference in its entirety) anti-ICOS mAb ( for example, MEDI-570, JTX-2011, and anti-TIM3 antibodies (see, for example, WO 2013/006490 or US Pat. prostate cancer, ovarian cancer, breast cancer, breast cancer, endometrial cancer, multiple myeloma, melanoma, lymphomas, lung cancer including lung cancer, kidney cancer, colorectal cancer, pancreatic cancer, stomach cancer, brain cancer (see usually www.clinicaltrials.gov).

3. Combination therapy for PVRIG and PD-1

As shown herein, the PVRIG antibodies of the invention can be combined with one of several checkpoint antibodies.

a) Anti-PVRIG antibodies in combination with anti-PD-1 antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and anti-PD-1 antibodies.

In one embodiment, the biopsy is taken from a tumor in a cancer patient and dissociated as known in the art for FACS analysis. Cells are stained with labeled antibodies to (1) PVRIG (generally using CHA.7.518.1H4 (S241P), for example, although any of these are described in WO 2016/134333 (in particular, any that bind even if they are not blocked) or WO 2017/041004); (2) PD-1 (eg, using those known in the art, including EH12.2H7, Keytruda®, Opdivo®, etc.); (3) PD-L1 (eg using those known in the art such as BM-1 described herein) and (4) PVRL2 (eg using those known in the art such as TX11); and (5) an isotype control antibody. FACS is done and for each receptor the percentage of cells expressing the receptor relative to the control antibody is calculated. If the percentage of positive cells for PVRIG, PD-1, PD-1 and PVRL2 is > 1% for all 4 receptors, then the patient is treated with antibodies to PVRIG and PD1 as described in this document.

There are two approved anti-PD-1 antibodies, pembrolizumab (Keytruda®) and nivolumab (Opdivo®) and many others in development, that can be used in combination with the anti-PVRIG antibodies of the invention.

Accordingly, the invention provides specific combinations of: CHA.7.518.1.H4 (S241P) (as depicted in Figure 3) with pembrolizumab; CHA.7.518.1.H4 (S241P) as shown in FIG. 3 with nivolumab; CHA.7.538.1.2.H4 (S241P) as shown in FIG. 3 with pembrolizumab and CHA.7.538.1.2.H4 (S241P) as shown with nivolumab.

- 47 040773

b) Anti-PVRIG antibodies in combination with aHTu-CTLA-4 antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and anti-CTLA-4 antibodies. There are two approved anti-CTLA-4 antibodies, ipilimumab (Yervoy®) and tremelimumab, as well as others in development, that can be used in combination with the anti-TIGIT antibodies of the invention.

Accordingly, the invention provides specific combinations of: CHA.7.518.1.H4 (S241P) with ipilimumab; CHA.7.518.1.H4 (S241P) with tremelimumab; CHA7.538.1.2.H4 (S241P) with ipilimumab and CHA7.538.1.2.H4 (S241P) with tremelimumab.

c) Anti-PVRIG antibodies in combination with anti-PD-L1 antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and anti-PD-L1 antibodies. There are three approved anti-PD-L1 antibodies, atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®) and durvalumab, as well as other anti-PD-L1 antibodies in development that can be used in combination with the anti-TIGIT antibodies of the invention.

Accordingly, the invention provides specific combinations of: CHA7.518.1.H4 (S241P) with atezolizumab; CPA7518.1.H4 (S241P) with avelumab; CHA.7.518.1.H4 (S241P) with durvalumab; CHA.7.538.1.2.H4 (S241P) with atezolizumab; CHA7.538.1.2.H4 (S241P) with avelumab and CHA.7.538.1.2.H4 (S241P) with duvalumab.

d) Anti-PVRIG antibodies in combination with Ahtu-LAG-3 antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and Ahtu-LAG-3 antibodies. Several Ahtu-LAG-3 antibodies are in development, including BMS-986016 (see International Patent Application No. WO 2010/019570A2, incorporated herein by reference) GSK2831781 (see US Patent Application No. 2016/0017037A, incorporated herein by reference) and clones Merck 22D2, 11C9, 4A10 and/or 19E8 (see WO 2016/028672A1, incorporated herein by reference in its entirety) and GSK2831781, as well as others in development that can be used in combination with anti-PVRIG antibodies according to the invention.

Accordingly, the invention provides specific combinations: CHA.7.518.1.H4(S241P) with BMS-986016; CHA,7.518.1.H4(S241P) with GSK2831781; CHA7.538.1.2.H4(S241P) c BMS-986016 and CHA.7.538.1.2.H4(S241P) c GSK2831781.

Accordingly, the invention also provides specific combinations of: CHA7.518.1.H4 (S241P) with Merck 22D2, 11C9 and/or 4A10 clones and CHA7.538.1.2.H4 (S241P) with Merck 22D2, 11C9 and/or 4A10 clones.

e) Anti-PVRIG antibodies in combination with anti-TIM-3 antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and anti-TIM-3 antibodies. There is at least one anti-TIM-3 antibody in development, TSR-022, as well as others in development that can be used in combination with anti-PVRIG antibodies of the invention.

Accordingly, the invention provides specific combinations: CHA7.518.1.H4(S241P) with TSR022 and CHA7.538.1.2.H4(S241P) with TSR-0226.

f) Anti-PVRIG antibodies in combination with anti-BTLA antibodies

In another embodiment, the invention provides combinations of anti-PVRIG antibodies of the invention and anti-BTLA antibodies, see WO2011/014438, incorporated by reference in its entirety, and in particular for the CDRs and full-length sequences of anti-BTLA antibodies described in her. Accordingly, the invention provides specific combinations of: CHA.7.518.1.H4 (S241P) with an anti-BTLA antibody and CHA.7.538.1.2.H4 (S241P) with an anti-BTLA antibody.

g) Anti-PVRIG antibodies with anti-tumor antibodies

In some embodiments, the anti-PVRIG antibodies of the invention are co-administered with antibodies that, unlike immuno-oncological/checkpoint inhibitors that typically act on the immune system to increase the patient's native immune response, are instead directed against a specific target tumor antigen (TTA ). There are a large number of anti-TTA antibodies approved or in development that can be combined with these anti-PVRIG antibodies, including CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P). Currently approved antibodies include, but are not limited to, cetuximab, panitumumab, nimotuzumab (all for EGFR), rituximab (CD20), trastuzumab and pertuzumab (HER2), alemtuzumab (CD52), bevacizumab (VEGF), ofatumumab (CD20), denosumab (RANK ligand), brentuximab (CD30), daratumumab (CD38), ibritumumab (CD20), and ipilimumab (CTLA-4). Specific targeted cancer antibodies in clinical trials that can be combined with anti-TIGIT antibodies include, but are not limited to, anti-CTLA4 mAbs such as ipilimumab, tremelimumab; anti-PD-1 such as nivolumab BMS-936558/MDX-1106/ONO-4538, AMP224, CT-011, MK-3475, anti-PDL-1 antagonists such as BMS-936559/MDX-1105, MEDI4736, RG- 7446/MPDL3280A; anti-LAG-3 such as IMP-321), anti-TIM-3, anti-BTLA, anti-B7-H4, anti-B7-H3, anti-VISTA; agonistic en

- 48 040773 titles targeting immunostimulatory proteins, including aHTu-CD40 mAbs, such as CP-870,893, lucatumumab, dacetuzumab; anti-CD137 such as BMS-663513 urelumab (anti-4-1BB, see, for example, US Pat. Nos. 7,288,638 and 8,962,804, incorporated herein by reference in its entirety); PF-05082566 utomilumab (see, for example, US Patent Nos. 8821867; 8337850; and 9468678 and International Patent Application Publication No. WO 2012/032433, incorporated herein by reference in its entirety); anti-OX40 mAbs, such as anti-OX40 (see, for example, WO 2006/029879 or WO 010096418, incorporated herein by reference in its entirety); anti-GITR mAbs such as TRX518 (see, for example, US Pat. No. 7,812,135, incorporated herein by reference in its entirety); anti-CD27 mAb, such as varlilumab CDX-1127 (see, for example, WO 2016/145085 and US Patent Publication No. US 2011/0274685 and US 2012/0213771, incorporated herein by reference in its entirety) anti-ICOS mAb ( e.g. MEDI-570, JTX2011 and anti-TIM3 antibodies (see e.g. WO 2013/006490 or US Pat. , ovarian cancer, breast cancer, breast cancer, endometrial cancer, multiple myeloma, melanoma, lymphomas, lung cancer including lung cancer, kidney cancer, colorectal cancer, pancreatic cancer, gastric cancer, brain cancer (see usually www.clinicaltrials.gov).

4. Combination therapy for PVRIG and TIGIT

There are specific combinations of anti-TIGIT and anti-PVRIG antibodies that find use in specific embodiments.

In one embodiment, the biopsy is taken from a tumor in a cancer patient and dissociated as known in the art for FACS analysis. Cells are stained with labeled antibodies to (1) PVRIG (typically using CHA.7.518.1H4(S241P), for example, although any of these are described in WO 2016/134333 (in particular, any that bind even if they are not blocked) or WO 2017/041004); (2) TIGIT (eg, using any of those described herein or others in the art, such as MBSA43); (3) PVR (eg using those known in the art such as SKII.4) and (4) PVRL2 (eg using those known in the art such as TX11); and (5) isotype control antibody. FACS is done and for each receptor the percentage of cells expressing the receptor relative to the control antibody is calculated. If the percentage of positive cells for PVRIG, TIGIT, PVR, and PVRL2 is > 1% for all 4 receptors, then the patient is treated with antibodies to PVRIG and TIGIT. Preferred combinations in this regard are CHA7.518.1.H4(S241P) and CPA.9.086.

In one embodiment, antibodies containing CDR sets from anti-TIGIT antibody CPA.9.086 are combined with antibodies containing CDR sets from anti-PVRIG antibody CHA.7.518.1. In a specific embodiment, antibodies containing the VH and VL sequences from anti-TIGIT antibody CPA.9.086 are combined with antibodies containing VL and VL from anti-PVRIG antibody CHA.7.518.1. In one embodiment, CPA.9.086.H4 (S241P), as shown in FIG. 53 is combined with CHA.7.518.1H4 (S241P) as shown in FIG. 3.

In one embodiment, antibodies containing sets of CDRs from anti-TIGIT antibody CPA.9.083 are combined with antibodies containing sets of CDRs from anti-PVRIG antibody CHA.7.518.1. In a specific embodiment, antibodies containing the VH and VL sequences from anti-TIGIT antibody CPA.9.083 are combined with antibodies containing VL and VL from anti-PVRIG antibody CHA.7.518.1. In one embodiment, CPA.9.086.H4(S241P) is combined with CHA7.518.1H4(S241P).

In one embodiment, antibodies containing CDR sets from anti-TIGIT antibody CPA.9.086 are combined with antibodies containing CDR sets from anti-PVRIG antibody CHA.7.538.1.2.H4(S241P). In a specific embodiment, antibodies containing the VH and VL sequences from anti-TIGIT antibody CPA.9.086 are combined with antibodies containing VL and VL from anti-PVRIG antibody CHA.7.538.1.2.H4(S241P). In one embodiment, CPA.9.086.H4(S241P) is combined with CHA7.538.1.2.H4(S241P).

In one embodiment, CHA.518.1.H4 (S241P) is combined with an anti-TIGIT antibody as specified in the sequence listing (with reference to all antibodies listed in FIG. 4 of USSN 62/513916), in particular

CPA.9.018, CPA.9.027,

CPA.9.049, CPA.9.057, CPA.9.059, CPA.9.083, CPA.9.086, CPA.9.089, CPA.9.093,

CPA.9.101, CPA.9.103, CHA.9.536.1, CHA.9.536.3, CHA.9.536.4, CHA.9.536.5, CHA.9.536.6,

CHA.9.536.7, CHA.9.536.8, CHA.9.560.1, CHA.9.560.3, CHA.9.560.4, CHA.9.560.5,

CHA.9.560.6, CHA.9.560.7, CHA.9.560.8, CHA.9.546.1, CHA.9.546.1, CHA.9.547.2,

CHA.9.547.3, CHA.9.547.4, CHA.9.547.6, CHA.9.547.7, CHA.9.547.8, CHA.9.547.9, CHA.9.547.13, CHA.9.541.1, CHA. 9.541.3. SNA.9.541.4. SHA.9.541.5, SHA.9.541.6.

SHA.9.541.7 and SHA.9.541.8

In one embodiment, CPA.9.086 is combined with an anti-PVRIG antibody as described in

- 49 040773

WO 2017/041004, including but not limited to those having: a) HC sequence SEQ ID NO: 5 and LC sequence SEQ ID NO: 3 (or CDR sets contained therein); b) the HC sequence of SEQ ID NO: 32 and the LC sequence of SEQ ID NO: 33 (or sets of CDRs contained therein); and c) the HC sequence of SEQ ID NO: 32 and the LC sequence of SEQ ID NO: 40 (or sets of CDRs contained therein).

In some embodiments, the combination comprises an anti-TIGIT antibody selected from the group consisting of CPA.9.086, CPA.9.083, CHA.9.547.7, and CHA.9.547.13, and an anti-PVRIG antibody selected from the group consisting of CHA7 .518.1 and CHA.7.538.1.2. In some embodiments, the combination comprises an anti-TIGIT antibody selected from the group consisting of CPA.9.086, CPA.9.083, CHA.9.547.7, and CHA.9.547.13, and the anti-PVRIG antibody is CHA7.518.1. In some embodiments, the combination comprises an anti-TIGIT antibody selected from the group consisting of CPA.9.086, CPA.9.083, CHA.9.547.7, and CHA.9.547.13, and the anti-PVRIG antibody is CHA7.538.1.2. In some embodiments, the combination comprises an anti-TIGIT CPA.9.086 antibody and an anti-PVRIG CHA7.518.1 antibody. In some embodiments, the combination comprises an anti-TIGIT CPA.9.083 antibody and an anti-PVRIG CHA7.518.1 antibody. In some embodiments, the combination comprises an anti-TIGIT antibody CHA.9.547.7 and an anti-PVRIG CHA7.518 antibody. In some embodiments, the combination comprises an anti-TIGIT CHA.9.547.13 antibody and an anti-PVRIG CHA7.518.1 antibody. In some embodiments, the combination comprises an anti-TIGIT CPA.9.086 antibody and an anti-PVRIG CHA7.538.1.2 antibody. In some embodiments, the combination comprises an anti-TIGIT antibody CPA.9.083 and an anti-PVRIG antibody CHA7.538.1.2. In some embodiments, the combination comprises an anti-TIGIT antibody CHA.9.547.7 and an anti-PVRIG antibody CHA7.538.1.2. In some embodiments, the combination comprises an anti-TIGIT antibody CHA.9.547.13 and an anti-PVRIG antibody CHA7.538.1.2.

In FIG. 20-24 are anti-PVRIG antibodies as described in U.S. Patent Application No. 15/277,978, filed September 27, 2016. The anti-TIGIT antibodies of this invention can be used in combination with anti-PVRIG antibodies as described in these figures, as well as with those disclosed in this application.

5. Treatment evaluation

Typically, antibodies of the invention, alone or in combination (PVRIG with PD-1, TIGIT with PD-1, or TIGIT with PVRIG), are administered to cancer patients and efficacy is assessed in several ways, as described herein. Thus, while standard efficacy analyzes such as tumor burden, tumor size, assessment of the presence or extent of metastasis, etc. can be performed, immuno-oncological treatments can also be evaluated based on assessments of immune status. This can be done in several ways, including both in vitro and in vivo assays. For example, changes in immune status can be assessed (eg, presence of ICOS+ CD4+ T cells after ipi treatment), along with legacy measurements such as tumor burden, size, invasiveness, lymph node involvement, metastasis, etc. Thus, any or all of the following can be assessed: inhibitory effects of PVRIG or TIGIT on CD4+ T cell activation or proliferation, CD8+ T cell (CTL) activation or proliferation, CD8+ T cell-mediated cytotoxic activity, and/or CTL-mediated cell population depletion , NK cell activity and NK-mediated cell population depletion, potentiating the effects of PVRIG or TIGIT on regulatory T cell differentiation and proliferation, and regulatory T cell or myeloid suppressor cell (MLSC)-mediated immunosuppression or immune tolerance, and/or the effects of PVRIG or TIGIT on the production of pro-inflammatory cytokines by immune cells, for example, the production of IL-2, IF H-γ or TNF-α with the help of T cells or other immune cells.

In some embodiments, treatment evaluation is performed by assessing immune cell proliferation using, for example, the CFSE dilution method, intracellular staining of Ki67 cells of immune effector cells, and the 3H-thymidine incorporation method.

In some embodiments, treatment evaluation is performed by assessing an increase in gene expression or an increase in protein levels associated with activation, including one or more of the following: CD25, CD69, CD137, ICOS, PD1, GITR, OX40, and cell degranulation as measured by CD surface expression. 107A.

In some embodiments, the treatment evaluation is performed by assessing the amount of T cell proliferation in the absence of treatment, for example, prior to administration of the antibodies of the invention. If, after administration, the patient has increased T cell proliferation, eg, a subset of the patient's T cells proliferate, this is indicative of T cell activation.

Likewise, evaluation of treatment with antibodies of the invention can be performed by measuring the patient's IFN-γ levels before and after administration to evaluate the effectiveness of the treatment. This can be done within hours or days.

In general, gene expression assays are performed as is known in the art (see, at

- 50 040773 example, Goodkind et al., Computers and Chem. Eng. 29(3):589 (2005), Han et al., Bioinform. Biol. Insights 11/15/15 9 (Suppl. l):29-46, Campo et al., Nod. Pathol. Jan 2013 26 suppl. 1:S97-S110, methods for measuring gene expression, which are expressly incorporated herein by reference.

In general, protein expression measurements are also performed in a similar manner as is known in the art, see, for example, Wang et al., Recent Advances in Capillary Electrophoresis-Based Proteomic Techniques for Biomarker Discovery, Methods. Mol. Biol. 2013:984:1-12; Taylor et al, BioMed Res. Volume 2014, Article ID 361590, 8 pages, Becerk et al., Mutat. Res 2011 June 17:722(2): 171-182, whose measurement methods are expressly incorporated by reference.

In some embodiments, treatment evaluation is performed by evaluating cytotoxic activity, measured by determining target cell viability by evaluating multiple cell parameters such as enzyme activity (including protease activity), cell membrane permeability, cell adhesion, ATP production, coenzyme production, and uptake activity. nucleotides. Specific examples of these assays include, but are not limited to, trypan blue or PI (propidium iodide) staining, ^51Cr or 35S release mode, lactate dehydrogenase (LDH) activity, MTT and/or WST assays, Calcein-AM assay, luminescence and others.

In some embodiments, treatment evaluation is performed by assessing T cell activity as measured by cytokine production, measured or intracellularly in culture supernatant using cytokines including, but not limited to, IFN-γ, TNF-α, GM-CSF, IL-2, IL -6, IL4, IL-5, IL-10, IL-13 using well known methods.

Accordingly, treatment can be evaluated using assays that evaluate one or more of the following: (i) increase in immune response, (ii) increase in αβ and/or γδ T cell activation, (iii) increase in T cell cytotoxic activity ( iv) increasing the activity of NK and/or NKT cells, (v) facilitating the suppression of αβ and/or γδ T cells, (vi) increasing the secretion of pro-inflammatory cytokines, (vii) increasing the secretion of IL-2; (viii) increase in interferon-γ production, (ix) increase in Th1 response, (x) decrease in Th2 response, (xi) decrease or kill cells and/or activity of at least one of the regulatory T cells (Treg).

Assays to measure effectiveness

In some embodiments, T cell activation is assessed using a mixed lymphocyte culture test (MLC test) as described in the examples. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in the immune response, as measured for example by phosphorylation or dephosphorylation of various factors, or by measuring other post-translational modifications. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in αβ and/or γδ T cell activation, as measured for example by cytokine secretion or proliferation, or changes in the expression of activation markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in cytotoxic T cell activity, as measured for example by direct killing of target cells, e.g., for cancer cells, or by cytokine secretion or by proliferation, or by changes in the expression of activation markers, such as, for example, CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in NK and/or NKT cell activity, as measured for example by direct killing of target cells, e.g., for cancer cells, or by cytokine secretion or by proliferation, or by changes in the expression of activation markers, as eg CD107a, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in αβ and/or γδ T cell suppression, as measured for example by cytokine secretion or by proliferation, or by changes in the expression of activation markers such as CD137, CD107a, PD1, etc. .d. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in secretion of pro-inflammatory cytokines as measured, for example, by ELISA or Luminex, or using a bead-based multiplex assay, or by intracellular staining and FACS assay, or Alispot, etc. . Increased activity indicates immunosuppression

- 51 040773 stimulating activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in IL-2 secretion as measured, for example, by ELISA or Luminex, or using a bead-based multiplex assay, or by intracellular staining and FACS assay, or Alispot, etc. . An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in interferon-γ production as measured, for example, by ELISA or Luminex, or using a bead-based multiplex assay, or by intracellular staining and FACS assay, or Alispot, etc. . An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in the Th1 response, as measured for example by cytokine secretion, or by changes in the expression of activation markers. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in the Th2 response, as measured for example by cytokine secretion, or by changes in the expression of activation markers. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in the number of cells and/or activity of at least one of the regulatory T cells (Treg), as measured, for example, by flow cytometry or IHC. Decreased response indicates immunostimulatory activity. The corresponding reductions are the same as for the increase described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in M2 macrophage cells, as measured, for example, by flow cytometry or IHC. Decreased response indicates immunostimulatory activity. The corresponding reductions are the same as for the increase described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in the protumor activity of M2 macrophages, as measured, for example, by cytokine secretion or by changes in the expression of activation markers. Decreased response indicates immunostimulatory activity. The corresponding reductions are the same as for the increase described below.

In one embodiment, the analysis of the signaling increases or decreases the increase in N2 neutrophils, as measured, for example, by flow cytometry or IHC. Decreased response indicates immunostimulatory activity. The corresponding reductions are the same as for the increase described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in N2 neutrophil protumor activity, as measured, for example, by cytokine secretion or by changes in the expression of activation markers. Decreased response indicates immunostimulatory activity. The corresponding abbreviations are the same as for magnification, described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in inhibition of T cell activation, as measured, for example, by cytokine secretion or by proliferation, or by changes in the expression of activation markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in CTL activation inhibition activity, as measured, for example, by direct killing of target cells, for example, for cancer cells, or by cytokine secretion, or by proliferation, or by changes in the expression of activation markers, as eg CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway analysis measures the increase or decrease in te and/or γδ depletion of T cells, as measured, for example, by changes in the expression of activation markers. Decreased response indicates immunostimulatory activity. The corresponding reductions are the same as for the increase, as shown below.

In one embodiment, the signaling pathway assay measures an increase or decrease in the response of αβ and/or γδ T cells, as measured, for example, by cytokine secretion or by proliferation, or by changes in the expression of activation markers such as CD137, CD107a, PD1, and etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in stimulation of antigen-specific memory responses, as measured by, for example, cytokine secretion or

- 52 040773 for proliferation, or for changes in the expression of activation markers such as CD45RA, CCR7, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in apoptosis or lysis of cancer cells, as measured for example by cytotoxicity assays, e.g., MTT, Cr release, Calcine AM, or flow cytometric assays, such as CFSE dilution or staining. propidium iodide, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in stimulation of a cytotoxic or cytostatic effect on cancer cells, as measured, for example, by cytotoxicity assays such as, for example, MTT, Cr release, Calcine AM, or flow cytometric assays, as, for example, CFSE dilution or propidium iodide staining, etc.

An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in direct killing of cancer cells, as measured, for example, by cytotoxicity assays, such as MTT, Cr release, Calcine AM, or flow cytometric assays, such as CFSE dilution or staining. propidium iodide, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in Th17 activity as measured, for example, by cytokine secretion or by proliferation or by changes in the expression of activation markers. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, the signaling pathway assay measures the increase or decrease in the induction of complement-dependent cytotoxicity and/or antibody-dependent cell-mediated cytotoxicity, as measured for example by cytotoxicity assays such as, for example, MTT, Cr release, Calcine AM, or by flow cytometry-based assay. such as CFSE dilution or propidium iodide staining, etc. An increase in activity indicates immunostimulatory activity. Corresponding increases in activity are described below.

In one embodiment, T cell activation is measured, for example, by direct killing of target cells, for example, for cancer cells, or by cytokine secretion or by proliferation, or by changes in the expression of activation markers, such as CD137, CD107a, PD1, etc. e. For T cells, increased proliferation, cell surface activation markers (eg, CD25, CD69, CD137, PD1), cytotoxicity (ability to kill target cells), and cytokine production (eg, IL-2, IL-4, IL -6, IFN-γ, TNF-α, IL-10, IL-17A) would be indicative of immune modulation consistent with increased cancer cell killing.

In one embodiment, NK cell activation is measured, for example, by direct killing of target cells, such as cancer cells or cytokine secretion, or changes in the expression of activation markers, such as for CD107a, etc. For NK -cell increase in proliferation, cytotoxicity (the ability to kill target cells and increase expression of CD107a, granzyme, and perforin), cytokine production (eg, IFN-γ and TNF), and cell surface receptor expression (eg, CD25) would be indicative of immune modulation that is consistent with increased destruction of cancer cells.

In one embodiment, γδ T cell activation is measured, for example, by cytokine secretion or proliferation or changes in the expression of activation markers.

In one embodiment, Th1 cell activation is measured, for example, by cytokine secretion or changes in the expression of activation markers.

The corresponding increase in activity or response (or decrease, if appropriate, as above) is an increase of at least about 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or from 98 to 99% of the signal in the standard or control samples, for example, test samples that do not contain anti-PVRIG antibodies according to the invention. Specific increases in activity are depicted in FIG. 27-34. For example, in terms of increasing T cell proliferation, CHA.7.518.1.H4 (S241P) shows an increase of about 60%, and CHA.7.538.1.2.H4 (S241P) shows an increase of 47%; corresponding increases are shown in both T cell proliferation and IFN-γ from about 10 to 70%, with 20 to 60% also finding use.

Likewise, an increase of at least one, two, three, four or five times compared to the standard or control samples has been shown to be effective.

XII. List of embodiments of the invention

1. Composition containing an antigen-binding domain that binds to human TIGIT (SEQ ID NO: 97), containing:

- 53 040773

a) a heavy chain variable domain comprising SEQ ID NO: 160; And

b) a light chain variable domain comprising SEQ ID NO: 165.

2. The composition according to claim 1, characterized in that said composition is an antibody containing:

a) a heavy chain containing a VH-CH1-hinge-CH2-CH3, said VH comprising SEQ ID NO: 160; And

b) a light chain containing VL-VC, said VL containing SEQ ID NO: 165, and VC is either kappa or lambda.

3. Composition according to claim 2, characterized in that the CH1-hinge-CH2-CH3 sequence is selected from human IgG1, IgG2 and IgG4 and variants thereof.

4. Composition according to claim 2 or 3, characterized in that said heavy chain has SEQ ID NO: 164 and said light chain has SEQ ID NO: 169.

5. A composition according to any one of claims 2 to 4, further comprising a second antibody that binds to a human checkpoint receptor protein.

6. The composition of claim 5, wherein said second antibody binds human PD-1.

7. The composition of claim 5, wherein said second antibody binds human PVRIG (SEQ ID NO: 2).

8. The composition according to claim 7, characterized in that said second antibody contains an antigen-binding domain containing a heavy chain variable domain containing SEQ ID NO: 5, and a light chain variable domain containing SEQ ID NO: 10.

9. The composition according to claim 7, characterized in that the heavy chain of said second antibody has SEQ ID NO: 9 and the light chain of said second antibody has SEQ ID NO: 14.

10. Nucleic acid composition containing:

a) a first nucleic acid encoding a heavy chain variable domain comprising SEQ ID NO: 160; And

b) a second nucleic acid encoding a light chain variable domain comprising SEQ ID NO: 165.

11. The composition of the nucleic acid according to claim 10, characterized in that the specified first nucleic acid encodes a heavy chain containing VH-CH1-hinge-CH2-CH3, while the specified VH contains SEQ ID NO: 160; and said second nucleic acid encodes a light chain comprising a VL-VC, wherein said VL comprises SEQ ID NO: 165 and the VC is a lambda domain.

12. An expression vector composition comprising a first expression vector containing said first nucleic acid according to claim 10 or 11 and a second expression vector containing said second nucleic acid according to claim 10 or 11, respectively.

13. An expression vector composition comprising an expression vector containing said first nucleic acid according to claim 10 or claim 11 and said second nucleic acid according to claim 10 or claim 11, respectively.

14. A host cell containing said expression vector composition according to claim 12 or 13.

15. A method for producing an anti-TIGIT antibody, comprising:

a) culturing said host cell according to claim 14 under conditions where said antibody is expressed; And

b) isolating said antibody.

16. A method of treating cancer by activating T cells, comprising administering a composition according to any one of claims 1-9.

17. Composition containing an antigen-binding domain that binds to human TIGIT (SEQ ID NO: 97), containing:

a) a heavy chain variable domain comprising SEQ ID NO: 150; And

b) a light chain variable domain comprising SEQ ID NO: 155.

18. The composition according to claim 17, characterized in that said composition is an antibody containing:

a) a heavy chain containing a VH-CH1-hinge-CH2-CH3, said VH comprising SEQ ID NO: 150; And

b) a light chain containing VL-VC, said VL containing SEQ ID NO: 159 and VC is either kappa or lambda.

19. Composition according to claim 18, characterized in that the CH1-hinge-CH2-CH3 sequence is selected from human IgG1, IgG2 and IgG4 and variants thereof.

20. A composition according to claim 17 or 18, characterized in that said heavy chain has SEQ ID NO: 154 and said light chain has SEQ ID NO: 159.

21. A composition according to any one of claims 17-20, further comprising a second antibody that binds to a human checkpoint receptor protein.

22. The composition of claim 21, wherein said second antibody binds human PD-1.

- 54 040773

23. The composition of claim 21, wherein said second antibody binds human PVRIG (SEQ ID NO:2).

24. The composition of claim 23, wherein said second antibody comprises an antigen-binding domain comprising a heavy chain variable domain comprising SEQ ID NO: 5 and a light chain variable domain comprising SEQ ID NO: 10.

25. The composition according to claim 23, characterized in that the heavy chain of said second antibody has SEQ ID NO: 9 and the light chain of said second antibody has SEQ ID NO: 14.

26. Nucleic acid composition containing:

a) a first nucleic acid encoding a heavy chain variable domain comprising SEQ ID NO: 150; And

b) a second nucleic acid encoding a light chain variable domain comprising SEQ ID NO: 155.

27. The composition of the nucleic acid according to item 26, characterized in that the specified first nucleic acid encodes a heavy chain containing VH-CH1-hinge-CH2-CH3, while the specified VH contains SEQ ID NO: 150; and said second nucleic acid encodes a light chain comprising a VL-VC, wherein said VL comprises SEQ ID NO: 155 and the VC is a lambda domain.

28. An expression vector composition comprising a first expression vector containing said first nucleic acid according to claim 26 or claim 27 and a second expression vector containing said second nucleic acid according to claim 26 or claim 27, respectively.

29. An expression vector composition comprising an expression vector containing said first nucleic acid according to claim 26 or claim 27 and said second nucleic acid according to claim 26 or claim 27, respectively.

30. A host cell containing said expression vector composition according to claim 27 or 28.

31. A method for producing anti-TIGIT antibodies, including:

a) culturing said host cell according to claim 30 under conditions where said antibody is expressed; And

b) isolating said antibody.

32. A method of treating cancer by activating T cells, comprising administering a composition according to any one of claims 17-25.

33. An antibody containing:

a) a heavy chain having SEQ ID NO: 9; And

b) a light chain having SEQ ID NO: 14.

34. The antibody of claim 33, further comprising a second antibody that binds to a human checkpoint receptor protein.

35. The composition of claim 34, wherein said second antibody binds human PD-1.

36. The composition of claim 34, wherein said second antibody binds human TIGIT (SEQ ID NO:97).

37. The composition of claim 36, wherein said second antibody comprises an antigen-binding domain comprising a heavy chain variable domain comprising SEQ ID NO: 160 and a light chain variable domain comprising SEQ ID NO: 165.

38. The composition of claim 36, wherein the heavy chain of said second antibody is SEQ ID NO: 164 and the light chain of said second antibody is SEQ ID NO: 169.

39. Nucleic acid composition containing:

a) a first nucleic acid encoding SEQ ID NO: 9; And

b) a second nucleic acid encoding SEQ ID NO: 14.

40. An expression vector composition comprising a first expression vector containing said first nucleic acid according to claim 39 and a second expression vector containing said second nucleic acid according to claim 39.

41. An expression vector composition comprising an expression vector containing said first nucleic acid according to claim 39 and said second nucleic acid according to claim 39.

42. A host cell containing said expression vector composition according to claim 41.

43. A method for producing an anti-PVRIG antibody, comprising:

a) culturing said host cell according to claim 42 under conditions where said antibody is expressed; And

b) isolating said antibody.

44. A method for treating cancer by activating T cells, comprising administering an antibody according to claim 33.

45. An antibody containing:

a) a heavy chain having SEQ ID NO: 19; And

b) a light chain having SEQ ID NO: 24.

46. The antibody of claim 45, further comprising a second antibody that binds to a human checkpoint receptor protein.

- 55 040773

47. The composition of claim 46, wherein said second antibody binds human PD-1.

48. The composition of claim 46, wherein said second antibody binds human TIGIT (SEQ ID NO: 97).

49. The composition of claim 48, wherein said second antibody comprises an antigen-binding domain comprising a heavy chain variable domain comprising SEQ ID NO: 160 and a light chain variable domain comprising SEQ ID NO: 165.

50. The composition of claim 49, wherein the heavy chain of said second antibody is SEQ ID NO: 164 and the light chain of said second antibody is SEQ ID NO: 169.

51. Nucleic acid composition containing:

a) a first nucleic acid encoding SEQ ID NO: 19; And

b) a second nucleic acid encoding SEQ ID NO: 24.

52. An expression vector composition comprising a first expression vector containing said first nucleic acid according to claim 51 and a second expression vector containing said second nucleic acid according to claim 51.

53. An expression vector composition comprising an expression vector containing said first nucleic acid according to claim 51 and said second nucleic acid according to claim 51.

54. A host cell containing said expression vector composition according to claim 53.

55. A method for producing an anti-TIGIT antibody, comprising:

a) culturing said host cell according to claim 54 under conditions where said antibody is expressed; And

b) isolating said antibody.

56. A method for treating cancer by activating T cells, comprising administering an antibody according to claim 45.

57. Method, including:

a) providing a population of cells from a tumor sample from the patient;

b) staining said population with labeled antibodies that bind:

i) TIGIT protein;

ii) PVR protein;

iii) PD-1 protein;

iv) PD-L1 protein and

v) appropriate isotype control for antibodies in i)-iv);

c) carrying out fluorescent activated cell sorting (FACS);

d) for each of TIGIT, PVR, PD-1 and PD-L1, determining the percentage of cells in said population that express the protein relative to said isotype control antibody;

however, if the percentage of positive cells is > 1% for all 4 receptors,

e) administering anti-TIGIT and PD-1 antibodies to said patient.

58. The method of claim 57 wherein said anti-TIGIT antibody is CPA.9.086.

59. The method of claim 57 or 58, wherein said anti-PD-1 antibody is selected from pembrolizumab and nivolumab.

60. Method, including:

a) providing a population of cells from a tumor sample from the patient;

b) staining said population with labeled antibodies that bind:

i) PVRIG protein;

ii) PVRL2 protein;

iii) PD-1 protein;

iv) PD-L1 protein and

v) appropriate isotype control for antibodies in i)-iv);

c) carrying out fluorescent activated cell sorting (FACS);

d) for each of PVRIG, PVRL2, PD-1 and PD-L1, determining the percentage of cells in said population that express the protein relative to said isotype control antibody;

however, if the percentage of positive cells is > 1% for all 4 receptors,

e) administering antibodies to PVRIG and PD-1 to said patient.

61. The method of claim 60 wherein said anti-PVRIG antibody is CHA.7.518.1.H4 (S241P).

62. The method of claim 60 or 61, wherein said anti-PD-1 antibody is selected from pembrolizumab and nivolumab.

63. Method, including:

a) providing a population of cells from a tumor sample from the patient;

- 56 040773

b) staining said population with labeled antibodies that bind:

i) PVRIG protein;

ii) PVRL2 protein;

iii) TIGIT protein;

iv) PVR protein and

v) isotype control;

c) carrying out fluorescent activated cell sorting (FACS);

d) for each of PVRIG, PVRL2, TIGIT and PVR, determining the percentage of cells in said population that express the protein relative to said isotype control antibody;

however, if the percentage of positive cells is > 1% for all 4 receptors,

e) administering antibodies to PVRIG and TIGIT to said patient.

64. The method of claim 63, wherein said anti-PVRIG antibody is CHA.7.518.1.H4 (S241P).

65. The method of claim 63 or 64, wherein said anti-TIGIT antibody is CPA9.086.

66. Composition containing an antigen-binding domain that binds to human TIGIT (SEQ ID NO: 97), containing:

a) heavy chain variable domain containing SEQ ID NO: 560; And

b) a light chain variable domain comprising SEQ ID NO: 565.

67. The composition according to claim 66, characterized in that said composition is an antibody containing:

a) a heavy chain containing a VH-CH1-hinge-CH2-CH3, said VH comprising SEQ ID NO: 560; And

b) a light chain containing VL-VC, said VL containing SEQ ID NO: 565 and VC is either kappa or lambda.

68. The composition according to claim 67, characterized in that the CH1-hinge-CH2-CH3 sequence is selected from human IgG1, IgG2 and IgG4 and variants thereof.

69. A composition according to claim 67 or 68, wherein said heavy chain has SEQ ID NO: 564 and said light chain has SEQ ID NO: 569.

70. A composition according to any one of claims 67-69, further comprising a second antibody that binds to a human checkpoint receptor protein.

71. The composition of claim 70, wherein said second antibody binds human PD-1.

72. The composition of claim 70, wherein said second antibody binds human PVRIG (SEQ ID NO:2).

73. The composition of claim 72, wherein said second antibody comprises an antigen-binding domain comprising a heavy chain variable domain comprising SEQ ID NO: 5 and a light chain variable domain comprising SEQ ID NO: 10.

74. The composition of claim 72, wherein the heavy chain of said second antibody is SEQ ID NO: 9 and the light chain of said second antibody is SEQ ID NO: 14.

75. Nucleic acid composition containing:

a) a first nucleic acid encoding a heavy chain variable domain comprising SEQ ID NO: 560; And

b) a second nucleic acid encoding a light chain variable domain comprising SEQ ID NO: 565.

76. The composition of the nucleic acid according to item 75, characterized in that the specified first nucleic acid encodes a heavy chain containing VH-CH1-hinge-CH2-CH3, while the specified VH contains SEQ ID NO: 560; and said second nucleic acid encodes a light chain comprising a VL-VC, wherein said VL comprises SEQ ID NO: 565 and the VC is a lambda domain.

77. An expression vector composition comprising a first expression vector containing said first nucleic acid according to claim 75 or 76 and a second expression vector containing said second nucleic acid according to claim 75 or 76, respectively.

78. An expression vector composition comprising an expression vector containing said first nucleic acid according to claim 75 or 76 and said second nucleic acid according to claim 75 or 76, respectively.

79. A host cell containing said expression vector composition according to claim 77 or 78.

80. A method for producing an anti-TIGIT antibody, comprising:

a) culturing said host cell according to claim 79 under conditions where said antibody is expressed; And

b) isolating said antibody.

81. A method of treating cancer by activating T cells, comprising administering a composition according to any

- 57 040773 from paragraphs 66-74.

XIII. Examples

Reference is made to PCT/US2016/18809, filed February 19, 2016, titled PVRIG ANTIBODIES AND METHODS OF TREATMENT, expressly incorporated herein by reference in its entirety and specifically to include examples 1-5, 7-8 , 11-13, 16-20 and 26-28, as well as the attached figures.

A. Example 1: PVR, PVRL2 and PVRL3 Binding Studies with PVRIG, DNAM and TIGIT by Surface Plasmon Resonance

Materials and methods

All experiments were performed using a ProteOn XPR 36 instrument at 22°C.

Stage 1:

A high density surface of a polyclonal anti-human fc antibody (Invitrogen H10500) was prepared in all six lanes of the GLC chip using a ProteOn XPR 36 biosensor. High-density PCAT occurred in the vertical flow direction. The blocking step took place in both vertical and horizontal positions so that horizontal intermediate points can be used as support surfaces. On average, an average of about 4400 RU of goat anti-human PCA was immobilized on each lane.

Stage 2:

For each run, three different lots of human PVRIG fusion protein (human fc, GenScript lots 451, 448, 125), human DNAM-1 fusion protein (human factors, R&D Systems), human TIGIT fusion protein (human factors, R&D Systems) and control Human IgG (Synagis) was captured on different vertical bands for two minutes at a concentration of 2 μg/ml. PVR, two batches of PVRL2 and PVRL3 were injected horizontally at six different concentrations across all six captured ligands in different ligand capture cycles. Injections lasted 2 min followed by 10 min dissociation at a flow rate of 50 µl/min. PVR concentration range was 1.4 nM-332 nM in 3-fold dilution series, both lots of PVRL2 were administered in the concentration range of 1.3 nM-322 nM in 3-fold dilution series and PVRL3 was administered in the concentration range of 1.4 nM-334 nM in 3-fold dilution series. All protein reagents were prepared in running buffer, which was degassed PBS buffer with 0.05% Tween 20 and 0.01% BSA. Anti-human fc capture surfaces were regenerated with two 30 second washes with 146 mM phosphoric acid after each cycle.

Stage 3:

The sensorgram data of analyte binding to each captured ligand was processed and presented under control using ProteOn Manager version 3.1.0.6 using reference to intermediate points and preliminary introduction of a working buffer identical to the introduction of the analyte.

results

a) PVR: binds weakly to captured DNAM-1 and TIGIT and shows no binding to all three batches of PVRIG and control IgG. Not enough information was collected to assess the KD interaction of PVR with DNAM-1 and TIGIT (data not shown).

b) PVRL2: Both lots of PVRL2 showed binding to all three lots of PVRIG and DNAM-1, but minimal or no binding to TIGIT and no binding to control IgG. The sensorgrams showed complex kinetics so binding constants could not be estimated (data not shown).

c) PVRL3: shows minimal binding to TIGIT and no binding to other proteins (data not shown).

B. Example 2: Effect of PVRIG Knockdown (KD) and Anti-FVRIG Antibodies on Human Melanoma-Specific TIL Functions

The purpose of these assays is to evaluate the functional capacity of PVRIG in TIL of human origin, as measured by markers of cytokine activation and secretion, when co-cultured with melanoma target cells.

1. Example 2(1):

The effect of an anti-PVRIG antibody (CPA.7.021), which has been shown to block the interaction of PVRIG and PVRL2, alone or in combination with other antibodies (eg, antiTIGIT, anti-DNAMl) was evaluated. PD1 was used as a reference immune checkpoint for specified knockdown (siRNA) studies.

Materials and methods:

TIL: Tumor infiltrating lymphocytes (TIL) from three melanoma patients were used (1) TIL-412-HLA-A2-Mart1, (2) TIL-F4-HLA-A2-gp100, and (3) TIL-209-HLA -A2-gp100. TIL was thawed in complete IMDM medium (BI, 01-058-1A) supplemented with 10% human serum (Sigma, H3667) + 1% glutamax (Life technologies, 35050-038) + 1% Na-pyruvate (Biological Industries, 03- 042-1B) + 1% replacement

- 58040773 non-amino acids (Biological Industries, 01-340-1B) + 1% Pen-Strep (penicillin/streptomycin) (Biological Industries, 03-031-1B) + 300 U/mL rhIL2 (Biolegend, 509129).

Tumor cell lines: Mel-624 human melanoma cells express MART-1 and gp-100 antigens in the MHC-I HLA-A2 haplotype. Cells were cultured in complete DMEM supplemented with 10%, 25 mM HEPES buffer, 1%, and 1% Pen-Strep.

Knockdown in TIL: Knockdown (KD) of human PVRIG and human PD1 in TIL was performed using 100 pmol Dharmacon ON-TARGETplus human PVRIG siRNA -SMARTpool (L-032703-02) or human PD1 siRNA -SMARTpool (L-004435)) or non-targeting siRNA (D-001810-01-05). siRNA was introduced by electroporation in TIL (AMAHA, program X-005). Electroporation was performed on resting TIL cultured in complete IMDM supplemented with IL-2 after 24 hours of thawing. After electroporation, TIL was subcultured into a 96-well TC plate for recovery within 24 h. After 24 h, cells were harvested and stained with viable score dye (BD Horizon, cat. no. 562247, BD biosciences), washed with PBS, and stained with anti-human PVRIG antibody - CPA.7.021 (CPA.7.021 IgG2 A647, 7.5 μg/ml) or anti-human PD-1 antibody (Biolegend, # 329910 AF647, 5 μg/ml) at room temperature for 30 min. The isotype controls used are synagis (IgG2 A647, 7.5 μg/ml) and mouse IgG1 (Biolegend No. 400130 A647, 5 μg/ml), respectively. All samples were measured on a MACSQuant analyzer (Miltenyi) and data was analyzed using FlowJo software (v10.0.8).

Co-cultivation of TIL with 624 melanoma cells:

TIL, electropolished siRNA, were collected and seeded in a 96-well TS-tablet 5x10 ⁴ /well. Mel-624 cells were also harvested and seeded in ratios of 1:1/1:3 E:T in co-culture. The tablet was incubated overnight (18 h) at 37°C, 5% CO ₂ .

To evaluate the effect of anti-PVRIG antibodies (CPA.7.021), anti-TIGIT (clone 10A7; from Genentech, US Patent Application No. US 2009/0258013) and anti-DNAMl (clone DX11, first described in Shibuya et al Immunity Vol. 4, Issue 6, 1 June 1996, Pages 573-581; BD Biosciences; mouse anti-human DNAM-1 clone DX11, Cat # 559787) for melanoma-specific TIL activity, TIL (1x105 cells/well) was preincubated with test antibodies or appropriate isotype controls in monotherapy (10 µg/ml) or in combination therapy (final concentration 10 µg/ml each) before adding 624 melanoma target cells at a 1:1 effector:target ratio. The tablet was incubated overnight (18 h) at 37°C, 5% CO ₂ .

Assessment of TIL activation: 16 h after co-culture, cells were stained with viability dye (BD Horizon, cat. no. 562247, BD biosciences), washed with PBS, and exposed to Fc-blocking solution (cat. no. 309804, Biolegend) followed by surface staining with anti -CD8a (cat. no. 301048, Biolegend) and anti-CD137 (cat. no. 309804, Biolegend) at 4°C for 30 min. All samples were run on a MACSQuant analyzer (Miltenyi) and data was analyzed using FlowJo software (v10.0.8). Culture supernatants were harvested and analyzed for cytokine secretion using a CBA kit (Cat. No. 560484, BD).

results

Knockdown of PVRIG in TIL: TIL MART-1 and TIL F4 were cultured for 24 h with IL-2. 100 pmol ONTARGETplus human PVRIG siRNA - SMART pool (L-032703-02) or human PD1 siRNA - SMARTpool (L-004435) or non-targeting siRNA (D-001810-01-05) were electroporated into TIL (AMAXA, program X- 005). Detection of PVRIG or PD-1 was performed 24 h after electroporation (and before co-culture). Cells were stained for viability followed by incubation for 30 min at room temperature with anti-PVRIG or anti-PD-1. The percentage of the KD population is indicated in FIG. 82 USSN 15/048967, incorporated herein by reference.

Functional analysis using knockdown TIL: Human TIL cultured for 24 h with IL-2 were electroporated with siRNA encoding human PVRIG or PD-1 or a scrambled sequence as a control. TILs were tested for PVRIG and PD-1 expression 24 hours after electroporation. ~80% PVRIG knockdown and ~50% PD-1 knockdown were observed compared to TIL electroporated with the scrambled sequence as depicted in FIG. 82 USSN 15/048967, incorporated herein by reference.

KD TIL were cultured with Mel-624 cells at 1:1 or 1:3 E:T for 18 h and stained for CD137 expression. Elevated levels of the activation marker CD137 were shown in TIL MART-1 electroporated with PVRIG siRNA, similar to TIL electroporated with PD-1 siRNA, compared to control scrambled siRNA (as depicted in FIG. 83A USSN 15/048967 included herein). document by reference). The co-culture supernatant was collected and tested for the presence of secreted cytokines. TILs that were electroporated with PVRIG siRNA show a significant increase in IFN-γ and TNF levels compared to control SCR siRNA (SCR-scrambled). A similar effect was demonstrated in TILs that were electroporated with PD-1 siRNA (as depicted in FIG. 83B-C USSN 15/048967, incorporated herein by reference).

- 59 040773

The same trend of increasing activation levels was observed in TIL F4. Co-culture of TIL F4 electroporated with PVRIG siRNAs with Mel-624 at 1:3 E:T resulted in increased levels of CD137 surface expression as well as increased IFN-γ secretion in the co-culture supernatant, as shown in FIG. 84A and 84B USSN 15/048967, incorporated herein by reference.

Similar trends were observed in TILs that were electroporated with PD-1 siRNA.

Functional analysis using blocking Abs:

In vitro monotherapy and combination therapy of anti-PVRIG and anti-TIGIT: 209 TIL were cultured with Mel-624 cells at 1:1 E:T for 18 h. Co-culture supernatant was harvested and tested for the presence of secreted cytokines. Anti-TIGIT treatment had no effect on IFN-γ or TNF secretion levels. However, an increase in IFN-γ and TNF levels was observed when anti-TIGIT and anti-PVRIG were added to co-culture in combination (Fig. 8A-B).

In vitro monotherapy and combination therapy of anti-PVRIG and anti-TIGIT: 209 TIL were cultured with Mel-624 cells at 1:1 E:T for 18 h. -DNAM-1. The co-culture supernatant was collected and tested for the presence of secreted cytokines. Treatment with anti-DNAM-1 mediated a trend towards increased secreted cytokines IFN-γ and TNF. Treatment with anti-DNAM-1 and anti-PVRIG in combination partially reversed the effect on CD137 expression (FIG. 9C) and enhanced the effect on IFN-γ and TNF cytokine secretion (FIG. 9A-B). MART-1 TIL was cultured with Mel-624 cells at 1:1 E:T for 18 h. Co-culture supernatant was harvested and tested for the presence of secreted cytokines. Treatment with anti-DNAM-1 reduced the surface expression of CD137 on TIL, as well as the secreted cytokines IFN-γ and TNF. Treatment with anti-DNAM-I and anti-PVRIG in combination partially reversed these effects (Fig. 9D-F).

SUMMARY OF THE INVENTION AND CONCLUSIONS: PD1 KD improves TIL activity as measured by IFN-γ and secretion in F4 and MART-1 TIL. An increase (~20%) in IFN-γ and TNF secretion was observed with PVRIG knockdown in MART-1 TIL compared to control siRNA. The same trend was observed for CD137 expression when co-cultivated with 624 HaF4 TIL melanoma cells.

Anti-TIGIT treatment did not affect IFN-γ or TNF secretion levels from TIL co-cultured with 624 Mel, however, an increase in IFN-γ and TNF levels was observed when anti-TIGIT and anti-PVRIG (CPA.7.021) were added to co-culture in combination .

Treatment with anti-DNAM-1 reduced the activation of TIL-MART-1, which is manifested by a decrease in the secretion of CD137 and cytokines, and anti-PVRIG (CPA.7.021) could partially reverse this effect in combination therapy with Anti-DNAM-1 Abs. In TIL 209, IFN-γ and TNF secretion levels were slightly elevated (~10%) with anti-DNAM-1, and increases in IFN-γ and TNF levels (~40 and 30% respectively) were observed when anti-DNAM1 and anti- PVRIG (CPA.7.021) was added to the co-culture in combination. Taken together, our results demonstrate that PVRIG is a novel co-inhibitory receptor for PVRL2.

2. Example 2 (2):

The effect of additional anti-PVRIG antibodies (CHA.7.518.1.H4 (S241P), CHA.7.524, CHA.7.530, CHA.7.538) was evaluated, which demonstrated blocking the interaction of PVRIG and PVRL2, alone or in combination with other antibodies (for example , anti-TIGIT, PD1), on the activity of TIL-209, TIL-412 and TIL-463-F4 when co-cultured with the 624 melanoma cell line.

The functional antibodies used in this assay were anti-hPVRIG hybridoma Abs (mIgG1 backbone) - CHA7.518.1.H4 (S241P); SNA.7.524; SHA.7.530; SHA.7.538 (lot M1 No. 30816); anti-hTIGIT (mIgG1 backbone) - clone 10A7 (Genescript), anti-TIGIT clone MBSA43 (e-biosciences) and mIgG1 (Cat. No. 400166, clone MORS-21, Biolegend)

Co-culture TIL and 624 Mel: TIL was thawed and cultured as described in 2.1 24 hours prior to co-culture. Tested Abs were added alone (10 μg/ml) or in combination with anti-TIGIT (20 μg/ml) to seeded TILs and incubated (total 100 μl) for 1 h at 37°C, 5% CO ₂ . Mel-624 cells were harvested and seeded at a 1:3 effector:target ratio in co-culture with TIL. The tablet was incubated overnight (18 h) at 37°C, 5% CO ₂ .

Assessment of TIL functional capacity: T cell activity was assessed based on the determination of IFN-γ in co-culture supernatants. Culture supernatants were harvested and tested for cytokines using the CBA kit (cat. no. 560484, BD) or the human IFN-γ/TNF-α MAGPIX kit. An unpaired two-tailed Student's t-test was calculated. P < 0.05 was considered statistically significant.

results

Functional analysis using TIL and melanoma cells in the presence of anti-PVRIG hybridoma Abs: human TIL cultured for 24 h with IL-2 were co-cultured with Mel-624 cells at 1:3 E:T for 18 h and were tested for cytokine secretion. In FIG. 31 describes a representative experiment of 5-6 performed. TIL were co-cultured

- 60 040773 with 624 melanoma cells in the presence of anti-TIGIT or anti-PVRIG Ab (blue), or a combination of anti-TIGIT and anti-PVRIG (green) and tested for IFN-γ/TNF secretion. In this experiment, all 4 anti-PVRIG Ab monotherapies increased (20-30%) IFN-γ secretion in 2 of 3 TILs tested (TIL209 and TIL463-F4), while in combination with anti-TIGIT all anti-PVRIG Abs CHA.7.518.1.H4 (S241P), CHA.7.530, CHA.7.538 increased IFN-γ secretion compared to antiTIGIT treatment alone.

The effect of Ab CHA.7.518.1.H4 (S241P) was statistically significant in experiments on TIL 463-F4gp100 in 5 experiments both alone and in combination with anti-TIGIT (Fig. 9E, G). The effect of anti-PVRIG Ab CHA.7.518.1.H4 (S241P) combination therapy was also statistically significant in TIL 209 (Fig. 9C). The effect of combination therapy with anti-PVRIG Ab CHA.7.538 was statistically significant in TIL 463-F4-gp100 (Fig. 9F).

Summary and conclusions:

In the experimental systems described herein, we observed the effect of anti-PVRIG on TIL in response to melanoma target cells, as seen from changes in IFN secretion. The tested anti-PVRIG hybridoma Abs mediated an increase in IFN-γ secretion compared to the corresponding isotype control. The CHA.7.518.1.H4 Ab (S241P) appears to have an advantage in mediating increased IFN-γ secretion as a monotherapy and over other tested aPVRIG Abs, however the magnitude of these effects varies between different TILs. This effect is enhanced in combination with anti-TIGIT treatment.

3. Example 2(3):

The aim is to evaluate the functional activity of anti-human PVRIG antibodies (CHA7.518.1.H4 (S241P), CHA.7.544 or CHA.7.538) on human TIL activity when co-cultured with CHO-S cells injected with a peptide stably co-expressing HLAA2, b2- microglo6ulin (B2M) and PVRL2.

We used TIL from resected metastases of three patients with melanoma: TIL-412-HLA-A2Mart1 (26-35) specific, TIL-463-F4-HLA-A2-gp100 (209-217) specific, TIL-463-F5- HLA- A2gp100 (209-217) specific and TIL-209-HLA-A2-gp100 (209-217) specific.

TIL was thawed in complete IMDM supplemented with 10% human serum + 1% glutamax + 1% Na-pyruvate + 1% non-essential amino acids + 1% Pen-Strep + 300 U/ml rhIL2 (Biolegend, 589106).

CHO-S cells (target cells) were stably transduced with a lentivirus expressing HLA-A2/B2M (lentiviral vector cat. no. CD515B-1-SBI, system biosciences) and grown under selection conditions with hygromycin B 600 μg/ml in CD CHO medium (p/n 10743-011) supplemented with 8 mM GlutaMax 1% and 1% Pen/Strep. Cells expressing HLA-A2/B2M were then cloned by dilution limit. Clone 3E8 highly expressing HLA-A2 and B2M was then transduced with a lentivirus expressing human PVRL2 (lentiviral vector cat. no. CD510B-1-SBI, system biosciences)) and grown under selection conditions with 6 μg/ml puromycin.

In the experimental system described herein (see FIG. 35), gp100 or MART-1 reactive TILs that endogenously express TIGIT, DNAM-1 and PVRIG, FIG. 37) were co-cultured with CHO-S HLA-A2/B2M/PVRL2 cells injected with the peptide.

The functional antibodies used in this assay were anti-human PVRIG; Ab 461 (Aldeveron) - referred to as 544 in this example, anti-human PVRIG chimeric Ab (hIgG4 backbone) - CHA.7.538; CHA.7.518 (referred to as c538 and c518 in this example, meaning that the heavy and light chain variable regions of 7.538 and 7.518 were fused to human IgG4 constant regions against human TIGIT (mIgG1 backbone), clone MBSA43 (e-biosciences), mIgG1 (biolegend) and hIgG4 (biolegend).

TIL was thawed and cultured as described herein for 24 hours prior to co-culture with target cells. Tested antibodies were added alone (10 μg/ml) or in combination with anti-TIGIT (total 20 μg/ml) to seeded TILs and incubated (total 100 μl) for 30 min at 37° C., 5% CO ₂ . CHO-S cells were harvested and injected with 0.1 or 0.5 μg/ml gp-100 (gp100209-217) or 20 μg/ml MART-126-35 peptides for 1 h at 37°C in serum reduced medium Opti-MEM™. After three washes with reduced Opti-MEM™ serum media, target cells injected with the peptide were co-cultured overnight (18 hours) with TIL at an effector:target ratio of 1:3 (33k:100k).

Evaluation of TIL performance: The effect of anti-PVRIG antibodies (10 μg/mL) as anti-TIGIT monotherapy or combination therapy on TIL activity was assessed by measuring cytokine secretion from co-culture supernatants that were cultured overnight using the Combined Bead Kit Array (CBA) (cat. no. 560484, BD). All samples were read on a MACSQuant analyzer (Miltenyi) and data was analyzed using FlowJo software (v10.0.8).

Dose-dependent response of anti-PVRIG antibodies: The effect of dose-dependent response of anti-PVRIG antibodies c518, c538 (or hIgG4 isotype control) was tested in the described assay with antibody concentrations of 30, 10, 3, 1, 0.3, 0.1 and 0.03 µg/ml. An unpaired two-tailed Student's t-test was calculated. P < 0.05 was considered statistically significant.

results

Effect of anti-PVRIG antibodies on TIL activity when co-cultured with PVRL2-expressing CHO-S HLA-A2/B2M cells: effect of three anti-PVRIG antibodies (544, c538 and c518) on the activity of four different TILs (412, 463), 462 and 209) from two different experiments are summarized in FIG. 37. Ab served as a control non-blocking Ab. Detailed experimental results are shown in Fig. 39. Treatment with antibodies 544, c538 and c518 increased the levels of secretion of IFN from TIL (by an average of 6%, 28% and 23%, respectively) compared with isotype antibody treatment. Increased IFN secretion was found in TIL treated with c538 or c518 compared to 544, a non-blocking control. No significant differences were found between treatments c538 to c518 At. Treatment with anti-TIGIT increased secretion of IFN from TIL (mean 49%) compared to the isotype. Combined treatment of c518 and c538 with anti-TIGIT induced an additive effect on IFN secretion from TIL, but the combined effect was not statistically significant compared to TIGIT alone.

Dose-Effect of Anti-PVRIG Antibodies on TIL Functionality: The effect of adding anti-PVRIG antibodies (c538 and c518) in a dose-dependent response on F4 and 209 TIL activity was evaluated (FIG. 80). The EC50 of the c518 and c538 antibodies is in the 1 digit nM compared to the isotype control as measured by the effect of TNF-α secretion from TIL.

Summary and Conclusions: In the experimental system described herein, we observed the effect of anti-PVRIG antibodies on TIL activity in response to co-culture with CHO-S target cells injected with the peptide, HLA-A2/B2M, expressing PVRL2. The anti-PVRIG antibodies that were tested mediated enhanced secretion of IFN and TNF from TIL compared to the corresponding isotype control. Antibodies c518 and c538 have a statistically significant advantage (p-0.0063 and p-0.0034, respectively) on TIL activity, which is manifested in the secretion of IFN, compared to 544, which is a non-blocking PVRIG antibody (based on a competitive experiment conducted by on cells expressing PVRIG). Both antibodies c518 and c538 had an additive effect with the anti-TIGIT antibody (statistically significant).

4. Example 2(4)

The purpose of this example was to evaluate the functional capacity of PVRIG in TIL of human origin, as measured by cytokine secretion, when co-cultured with melanoma target cells. The effect of anti-PVRIG antibodies (CHA.7.518.1.H4 (S241P), CHA.7.524, CHA.7.530, CHA.7.538) was evaluated, which demonstrated blocking the interaction of PVRIG and PVRL2 alone or in combination with other antibodies (for example, anti -TIGIT, PD1).

Purified CD3+ T lymphocytes were obtained using a human T cell enrichment kit (Stem cell technologies) on blood samples to obtain a buffy coat. Cells were thawed and labeled with CFSE (Moleculare probes) to monitor proliferation in co-culture.

CHO-S-OKT3 Cells: CHO-S cells were transduced with CD5L-OKT3-scFv-CD14 into CD710B-1 (SBI Cat # CS965A-1, Lot # 151014-005, 1.40 x 10 ⁸ ifus/ml). Cells were cultured in the presence of CD CHO (Gibco, life technologies, cat. no. 10743-011) supplemented with 8 mM glutamax and 6 μg/ml puromycin. Surface levels of OKT3 were assessed by flow cytometry using RE goat anti-mouse IgG F(ab)'2 at a 1:200 dilution (Jackson Immunoresearch cat. no. 115-116-146). CHO-S-OKT3 cells were then transiently transfected with human PVRL2 (delta isoform) or empty vector using the Amaxa electroporation system (Lonza, Walkersville, MD, USA) according to the manufacturer's instructions. 5 μg of pcDNA3.1 plasmid (empty vector or hPVRL2) in 2x10 ⁶ cells in Ingenio™ electroporation solution (Minis, cat. no. MC-MIR50115) and pulse program U-024 were used. PVRL2 expression on transfected CHOS-S-OKT3 cells was assessed by flow cytometry using anti-PVRL2 Ab (cat. no. 337412, Biolegend).

The functional antibodies used in this assay were anti-hPVRIG hybridoma Abs (mIgG1 backbone) - CHA7.518.1.H4 (S241P); SNA.7.524; SHA.7.530; CHA.7.538, anti-TIGIT clone MBSA43 (e-biosciences) and mIgG1 (cat. no. 400166, clone MOPC-21, Biolegend).

Co-culture of CD3 cells and CHO-OKTS cells: CD3+ T cells were thawed and immediately labeled with CFSE. In parallel, CHO-S-OKT3-PVRL2 cells were harvested and treated with mitomycin-C for 1 hour at 37°C, washed and added to co-culture with T cells at 1:5 E:T (1 x 105 T cells and 2 x 10 ⁴ CHO-OKT3-PVRL2 or blank). Ab was added alone (10 μg/ml) or in combination with anti-TIGIT (10 μg/ml) and co-culture plates were incubated at 37° C., 5% CO 2 for 5 days. After 5 days, cells were harvested and T cell proliferation was analyzed by FACS gated on CD4 and CD8 subpopulations.

Effect of anti-PVRIG antibodies in CHOS-OKT3 co-culture assay: CFSE-labeled T cells were stimulated with stimulator cells (CHO cells expressing membrane-bound

- 62 040773 fragments of aHTu-CD3 mAb). CHOS stimulator cells expressing human PVRL2 and control stimulator cells (empty vector) treated with mitomycin C (50 μg/ml for 1 h) before co-culture with CFSE-labeled human T cells at a ratio of 1:5. After 5 days at 37°C and 5.0% CO ₂ the effect of anti-PVRIG antibodies (10 μg/ml) on T cell proliferation (CFSE dilution) and cytokine secretion (ELISA kits or TH1/2/17 CBA kits) in supernatants culture has been evaluated. All samples were read on a MACSQuant analyzer (Miltenyi) and data was analyzed using Flow Jo software (v 10.0.8). Culture supernatants were harvested and analyzed for cytokine secretion using a CBA kit (Cat. No. 560484, BD).

results

Effect of anti-PVRIG antibodies on PVRL2 overexpression in the CHOS-OKT3 assay: CHOSOKT3 cells overexpressing PVRL2 or control cells (empty vectors) were co-cultured with CD3+ cells, and the effect of anti-PVRIG antibodies as monotherapy or in combination with anti-TIGIT by T cell proliferation and cytokine secretion (FIG. 40). After 5 days, cells were harvested and analyzed for CFSE dilution. In parallel, the co-culture supernatant was harvested and tested for cytokine secretion. in fig. 41 shows the effect of anti-PVRIG Abs in a responding versus non-responding donor. The effect of various anti-PVRIG Abs on T-cell proliferation was evaluated as monotherapy in combination with anti-TIGIT. Although some anti-PVRIG Abs enhance T cell proliferation, no additive effect was observed with the anti-TIGIT antibody in this system (FIG. 42). These effects were not observed when testing for Abs co-cultured CD3+ cells with control (empty vector transfected) CHO-S cells (data not shown).

A total of 10 donors were tested and 5 out of 10 donors responded to anti-PVRIG Ab. CHA.7.518.1.H4(S241P) Ab therapy consistently resulted in an increase in IFN-γ secretion ranging from 20-50% in 5 responding donors tested, while treatment with other Abs did not show a clear trend (Fig. 43) . Similar effects were observed during the proliferation of CD8+ cells. The effect of Ab therapy is shown in FIG. 44.

C. Example 3: Effect of anti-PVRIG antibody on human melanoma-specific TIL functions in combination with anti-TIGIT and anti-PD1 antibodies

1. Example 3(1):

Materials and methods

TIL: Tumor infiltrating lymphocytes (TIL) from three melanoma patients were used (1) TIL-412-HLA-A2-Mart1 specific, (2) TIL-F4-HLA-A2-gp100 specific, and (3) TIL209-HLA -A2-gp100 specific.

TIL was thawed in complete IMDM medium (BI, 01-058-1A) supplemented with 10% human serum (Sigma, H3667) + 1% glutamax (Life technologies, 35050-038) + 1% Na-pyruvate (Biological Industries, 03042- 1B) + 1% non-essential amino acids (Biological Industries, 01-340-1B) + 1% Pen-Strep (penicillin/streptomycin) (Biological Industries, 03-031-1B) + 300 U/ml rhIL2 (Biolegend, 509129).

Tumor cell lines: Mel-624 human melanoma cells express MART-1 and gp-100 antigens in the MHC-I HLA-A2 haplotype. Cells were cultured in complete DMEM medium (Biological Industries, 01-055-1A) supplemented with 10% FBS (BI, 04-127-1 A), 25 mM HEPES buffer (BI, 03-025-1B), 1% glutamax ( Life Technologies, 35050-038) and 1% Pen-Strep (Biological Industries, 03-031-1B).

Co-culture of TIL with 624 melanoma cells in the presence of anti-PVRIG, anti-TIGIT and PD1 blocking antibodies: to assess the effect of anti-PVRIG antibodies (CPA.7.021), anti-TIGIT (clone 10A7) and anti-PD1 (mAb 1B8, Merck ) for melanoma-specific TIL activity, TIL (3x10 ⁴ cells/well) was pre-incubated with test antibodies or appropriate isotype controls in monotherapy (10 µg/ml) or in combination therapy (final 10 µg/ml each) before cells were added- melanoma targets 624 in a ratio of 1:3 effector: target. The plate was incubated overnight (18 hours) at 37°C, 5% CO2.

Assessing TIL activation: Culture supernatants were harvested and analyzed for cytokine secretion using a CBA kit (Cat. No. 560484, BD).

In vitro anti-PVRIG monotherapy and combination therapy with anti-TIGIT and PD1 blocking antibodies: F4 TIL (gp100 sepecific) were cultured with Mel-624 cells at 1:3 E:T for 18 h. Co-culture supernatant was harvested and tested for the presence secreted cytokines. Anti-TIGIT or anti-PD1 therapy had no effect on IFN-γ or TNF secretion levels. However, an increase in IFN-γ and TNF levels was observed when anti-TIGIT and anti-PVRIG or anti-PD1 in combination were added to the co-culture in combination (Fig. 10A-B).

Anti-PVRIG, anti-TIGIT, and PD1 therapy alone had no effect on IFN-γ or TNF secretion levels from TIL co-cultivation with 624 Mel, however, an increase in IFN-γ and TNF levels was observed when anti-TIGIT or anti-PD1 antibodies were added in combination with anti-PVRIG (CPA.7.021). The data presented indicate a synergistic effect for combination therapy with anti-TIGIT or anti-PD1 antibodies.

- 63 040773

2. Example 3(2):

Again, the ability of anti-PVRIG antibodies to enhance CD4+ and CD8+ T cell function was evaluated in combination with anti-TIGIT antibodies in a primary cell-based in vitro assay.

CHO-S OKT3: The CHO-S OKT3 assay was used to determine whether the combination of a humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) and a commercially available anti-TIGIT antibody, could increase T cell proliferation and cytokine secretion. more than therapy with one anti-PVRIG or anti-TIGIT. The target cells used in the co-culture assay were the Chinese hamster ovary cell line, CHO-S (ATCC), stably overexpressing the single chain variable fragment of the anti-human CD3 antibody clone OKT3 (abbreviated OKT3) and human PVRL2 (abbreviated hPVRL2). Stock CHO-S OKT3 cells were grown in serum-free CD-CHO medium supplemented with 40 mM glutamax, penicillin/streptomycin, and 6 μg/ml puromycin. CHO-S OKT3 hPVRL2 cells were grown in serum-free CD-CHO medium supplemented with 40 mM glutamax, penicillin/streptomycin, 6 µg/ml puromycin, and 600 µg/ml hygromycin B.

Primary CD3+ and CD8+ T cells were isolated from healthy human donors using RosentSep™ enriched mix for human CD3+ T cells (Stemcell Technologies) and human CD8+ microbeads (Miltenyi Biotec), respectively, and frozen in liquid nitrogen. On the day of the CD3+ or CD8+ co-culture assay, T cells were thawed, counted, and labeled with 1 μM CFSE (Life Technologies) for 10 min at 37°C. After this incubation, T cells were washed and resuspended in complete medium containing RPMI supplemented with 10% heat-inactivated FBS, (fetal bovine serum), glutamax, penicillin/streptomycin, non-essential amino acids, sodium pyruvate, and 50 μM □-mercaptoethanol. CHO-S OKT3 hPVRL2 cells were harvested from culture and treated with mitomycin C for 1 h at 37° C. with occasional agitation. After incubation, target cells were thoroughly washed, counted and resuspended in complete RPMI medium. The assay was performed with a 5:1 ratio of T cells (100,000) to target cells (20,000). Target cells, T cells, and antibody therapy at 10 μg/mL were added together in a 96-well U-bottomed plate (Costar) and incubated for 3 days (CD8+ T cells) or 5 days (CD4+ T- cells) at 37°C. Antibody-based therapies included human CHA.7.518.1.H4 (S241P) IgG4 alone, human IgG4 isotype control in combination with mouse anti-human TIGIT (clone MBSA43, eBioscience) and a combination of CHA.7.518.1.H4 (S241P) and anti- TIGIT (clone MBSA43). In addition, mouse IgG1 activity against human DNAM-1 (clone DX11, BioLegend), isotype control of mouse IgG1 (clone MOPC21, BioLegend) and isotype control of human IgG4 were also assessed.

After a 3 or 5 day incubation period, co-culture supernatants were analyzed for secreted cytokines, including IL-2, IL-4, IL-5, IL-6, IL-9, IL-10, IL-13, IL-17A, IL -17B, IL21, IL-22, TNF-α and IFN-γ, with a kit for the determination of human Th1/Th2/Th17 cytokines (BD Biosciences) by cytometric granular analysis (CBA) or with a LEGENDplex™ Human Th cytokine kit (BioLegend ). T cell proliferation was measured by staining CD4+ or CD8+ T cells with the LIVE/DEAD Dead Cell Aqueous Permanent Staining Kit (ThermoFisher Scientific) anti-CD4 antibody (clone RPA-T4, BioLegend) and anti-CD8 antibody (clone HIT8a, BioLegend ) and gated by percentage of live, CFSE-proliferating CD4+ or CD8+ T cells. Data were acquired using FACS Canto II (BD Biosciences) and analyzed using FlowJo (Treestar) and Prism (Graphpad) software.

Results: Combination of CHA.7.518.1.H4 (S241P) and anti-TIGIT antibody increases CD4+ T cell proliferation compared to antibody therapy alone: ability of humanized hybridoma-derived CHA.7.518.1.H4 (S241P) anti-PVRIG antibody to enhance proliferation of primary CD4+ T cells in vitro in combination with an anti-TIGIT antibody was assessed using the CHO-S OKT3 assay.

In FIG. 33A and B depict the percentage of proliferating CD4+ T cells from two different donors in response to co-culture with CHO-S OKT3 hPVRL2 target cells and treated with anti-PVRIG and anti-TIGIT antibodies, either alone or in combination. In these two exemplary human CD3+ T cell donors, the combination of CHA.7.518.1.H4 (S241P) and anti-TIGIT antibody increases CD4+ T cell proliferation compared to CHA.7.518.1.H4 (S241P) alone, or isotype combination IgG4 and anti-TIGIT antibodies. Anti-DNAM-I antibody reduces CD4+ T cell proliferation compared to IgG1 isotype control in both donors.

CHA.7.518.1.H4 (S241P) and anti-TIGIT antibody enhances CD8+ T cell proliferation and IFN-gamma secretion in FIG. 34A illustrates the ability of a humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P), to increase CD8+ T cell proliferation in combination with an anti-TIGIT antibody in the CHO-S OKT3 assay. In an exemplary human CD8+ T cell donor, the combination of CHA.7.518.1.H4 (S241P) and an anti-TIGIT antibody increases CD8+ T cell proliferation when the T cells are co-cultured with CHO-S OKT3 hPVRL2 cells. The combination of anti-PVRIG and antiTIGIT antibodies increased proliferation more than CHA.7.518.1.H4 (S241P) or treatment with the hIgG4 isotype plus anti-TIGIT antibody. In FIG. 34B shows that the same illustrative donor

- 64 040773

Human CD8+ T cells as described above, a humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P), in combination with an anti-TIGIT antibody also enhances IFN-γ secretion in the CHO-S OKT3 assay. The combination of anti-PVRIG and anti-TIGIT antibodies increases IFN-γ secretion more than CHA.7.518.1.H4 (S241P) or treatment with the hIgG4 isotype plus anti-TIGIT. The anti-DNAM-I antibody reduces both CD8+ T cell proliferation and IFN-γ production compared to an IgG1 isotype control antibody.

Summary and conclusions

Together, the humanized anti-PVRIG antibody, CHA.7.518.1.H4 (S241P) and the anti-TIGIT antibody had in vitro functional activity in the primary CHO-S OKT3 cellular assay. The combination of CHA.7.518.1.H4 (S241P) and anti-TIGIT antibody resulted in increased proliferation of CD4+ and CD8+ T cells, as well as secretion of IFN-γ from CD8+ T cells compared with therapy or CHA.7.518.1.H4 (S241P ), or anti-TIGIT antibody alone. Together, these data demonstrate that co-blockade of two checkpoints, PVRIG and TIGIT, increases T cell function compared to blockade of a single receptor.

It should be noted that TIGIT does not interact with CD112 (PVRL2, see Figures 4E and 4F Zhu et al, J. Exp. Med. (2016): 1-10); rather, it interacts with PVR, another ligand. PVR is expressed in the CHO/CD112 system by Zhu et al. Accordingly, our interpretation of the combined effect of aCD112R (anti-PVRIG antibody) and anti-TIGIT is that aCD112R/aPVRIG blocks interaction of human CD112R with human CD112, but anti-TIGIT antibody blocks interaction of human TIGIT with human or hamster PVR (on T cells). or CHO cells), Zhu et al. do not really provide a hypothesis as to why the combined Ahtu-CD112R/Ahtu-TIGIT effect occurs in the CD112 CHO assay. That is, the combined effect does not occur through the PVRL2/CD112 ligand alone.

D. Example 4: Epitope mapping of human anti-PVRIG antibodies based on cynomolgus monkey cross-reactivity

Rationale and goals

The aim of this study is to identify epitopes on the PVRIG protein that define the cross-reactivity of human PVRIG antibodies against the cynomolgus monkey (supo) orthologue. Many of the antibodies against the human PVRIG target show varying degrees of cross-reactivity with the cynomolgus monkey, despite the fact that many of these antibodies belong to the same epitope group. To shed light on the molecular basis of human/cynomolgus cross-reactivity (or lack of it), several cynomolgus to human mutations of recombinant PVRIG proteins were constructed, expressed and purified and tested for binding to a panel of anti-human PVRIG antibodies in ELISA.

Ways

Construction of PVRIG variants of cynomolgus-to-human:

Human sequence alignment and PVRIG EVA show 90% sequence identity and 93% sequence homology between human and cynomolgus monkey orthologues. Based on the nature of the mutations (conservative vs. non-conservative) and the prediction of the secondary structure (helix vs. extended) of the mutation region, three site-directed cynomolgus macaque PVRIG mutants were designed to map epitopes directed towards cynomolgus macaque cross-reactivity. These mutants include H61R, P67S and L95R/T97I PVRIG cynomolgus monkey. Wild-type cynomolgus macaque and human PVRIG have also been generated.

Expression and purification of cynomolgus monkey, human and hybrid PVRIG variants: All PVRIG variants were expressed as hybrids of VKD with a C-terminal 6XHis tag in mammalian cells. Proteins were purified using affinity purification, ion exchange chromatography and size exclusion chromatography. Purified proteins were dialyzed against PBS buffer (pH 7.4) and stored at 4°C.

ELISA to determine PVRIG-antibody interaction: Functional ELISA was performed as follows: recombinant PVRIG (His-labeled) cynomolgus, human, and cynomolgus/human hybrid proteins were adsorbed onto an IA plate overnight at 4°C. Coated plate wells were washed twice with PBS and incubated with 300 μl of blocking buffer (5% skimmed milk powder in PBS pH 7.4) at room temperature (RT) for 1 hour. Blocking buffer was removed and the plates were washed twice with PBS. Plate-bound PVRIG variants were incubated with anti-human PVRIG mAb (human IgG1 isotype) in solution (linear range 0.1 µg/mL to 8 µg/mL in 50 µL/well) at room temperature for 1 h. The plates were washed three times with PBS-T (PBS 7.4, 0.05% Tween20), then PBS and 50 μl/well HRP (horseradish peroxidase) conjugated secondary antibody (specific for human Fc IgG, Jackson ImmunoResearch) were added three times. This was incubated at room temperature for 1 h and the plates were washed again. ELISA signals were developed in all wells by adding 50 µl of Sureblue TMB substrate (KPL Inc) and incubating for 5-20 min. The reaction with HRP was stopped by adding 50 µl of 2N H2SO4 (VWR) and the absorbance signals at 450 nm were read on a SpectraMax spectrophotometer (Molecular Devices) or

- 65 040773

EnVision (PerkinElmer). Data was exported to Excel (Microsoft) and plots were built in GraphPad Prism (GraphPad Software, Inc.).

results

S67, R95 and 197 as cynomolgus monkey cross-reactivity determinants: binding data shown in FIG. 18 clearly demonstrate that residues S67, R95 and 197 affect the cross-reactivity of the cynomolgus monkey with different antibodies. While the cynomolgus-to-human mutation P67S negatively affects the binding of CPA.7.002 and CPA.7.041, the cynomolgus-to-human mutation L95R/T97I significantly improves the binding of CPA.7.002, CPA.7.021, CPA.7.028 and CPA. 7.041. On the other hand, the cynomolgus-to-human H61R mutation does not affect the binding of any of the tested antibodies.

Relative binding to cynomolgus-to-human variants suggests three epitope groups: relative antibody binding to cynomolgus, human and hybrid PVRIG variants suggests 3 different epitope groups: group 1 binds to residues R95/I97 (CpA.7.021 and CPA.7.028 ). Group 2 binds to residues S67 and R95/I97 (CPA.7.002 and CPA.7.041). Group 3 does not bind to residues S67 or R95/I97 (CPA.7.024 and CPA.7.050). The epitope groups show a strong correlation with the degree of cynomolgus monkey cross-reactivity of these antibodies (FIG. 19).

Summary and Conclusions: Limited epitope mapping based on human cynomolgus macaque variation in PVRIG EVA identified residues S67, R95, and 197 as determinants of cynomolgus macaque anti-human PVRIG antibody cross-reactivity. Complete restoration of binding to L95R/T97I of cynomolgus monkey PVRIG for antibodies CPA.7.021 and CPA.7.028 and improved binding of CPA.7.002 to this mutant strongly indicates that residues R95 and 197 are critical human PVRIG epitopes for these antibodies. These data also point to a possible way to predict the cross-reactivity of non-human primate PVRIG orthologues based on their primary amino acid sequence.

E. Example 5: Humanized Antibodies: Binding and Receptor-Ligand Blocking Assay of Humanized Anti-PVRIG Antibodies of Hybridoma Origin, CHA.7.518.1 .H4(S241P)HCHA.7.538.1.2.H4(S241P)

This experiment was performed to characterize the binding of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) to human and cynomolgus monkey PVRIG protein on cell lines and primary leukocytes to characterize the ability of CHA.7.518.1. H4 (S241P) and CHA.7.538.1.2.H4 (S241P) to block the interaction between PVRIG and PVRL2 and characterize the epitope placement of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) relative to each other another, by evaluating competition for binding to the PVRIG antigen expressed on Jurkat cells.

FACS analysis of hPVRIG overexpressing cells: The following cell lines were used to evaluate the specificity of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P): parental HEK and HEK cells overexpressing cells. These cells were cultured in DMEM (Gibco) + 10% fetal calf serum (Gibco) + glutamax (Gibco). For HEK cells overexpressing hPVRIG, 0.5 μg/ml puromycin (Gibco) was added to the positive selection medium. For FACS analysis, all cell lines were harvested in logarithmic growth phase and 50,000-100,000 cells per well were plated in 96-well plates. The binding of unconjugated CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (hIgG4) and their respective controls were evaluated in 8-point titration series starting at 10 μg/mL on ice for 30 min - 1 h Series titration was carried out in the form of three-fold serial dilutions. Unconjugated primary antibodies were detected using an anti-human Fc antibody conjugated to Alexa 647 (Jackson Laboratories). Data were acquired using FACS Canto II (BD Biosciences), FACS LSR Fortessa X-20 (BD Biosciences) or IntelliCyt (IntelliCyt Corporation) and analyzed using Flow Jo (Treestar) and Prism (Graphpad) software.

FACS analysis of human cell lines for hPVRIG: The following cell lines were used to evaluate the expression and specificity of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P): Jurkat and HepG2. Jurkat cells were cultured in RPMI + 10% fetal calf serum, glutamax, essential amino acids (Gibco), sodium pyruvate (Gibco), and penicillin/streptomycin (Gibco). HepG2 cells were cultured in DMEM + 10% fetal calf serum + glutamax. For FACS analysis, all cell lines were harvested in logarithmic growth phase and 50,000-100,000 cells per well were plated in 96-well plates. The binding of unconjugated CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (hIgG4) and their respective controls were assessed in an 8-point titration series starting at 10 μg/mL on ice for 30 min - 1 h. Unconjugated primary antibodies were detected using an anti-human Fc antibody conjugated to Alexa 647. The titration series was performed as 3-fold serial dilutions. Data was acquired using FACS Canto II or IntelliCyte and analyzed using FlowJo and Prism software.

- 66 040773

FACS PVRIG analysis on expanded CMV CD8 T cells: CMV reactive donors were purchased from Cellular Technology Limited (CTL). The supplied PBMCs were injected with 10 μM CMV peptide 494-503 (NLVPMVATV, Anaspec) for 2 hours. PBMCs were then washed three times before being plated in 24-well plates for 9 days in RPMI + 10% human AV serum (Sigma), glutamax, penicillin/streptomycin and cytokine growth mix consisting of 2 ng/mL IL-2 (R&D systems) and 10 ng/ml IL-7 (R&D systems). After 9 days, non-adherent cells were harvested, phenotyped to enrich CD8 T cells, and stored in liquid nitrogen.

To assess expression on expanded CMV CD8 T cells, the binding of unconjugated CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) (hIgG4) and their respective controls were assessed using 8-point titration series starting at 666 nM on ice for 30 min - 1 h. Titration series were performed as 4-fold dilution series. Unconjugated primary antibodies were detected using anti-human Fc antibody conjugated to Alexa 647. Data were analyzed using FlowJo and Prism software and collected on a BD LSR Fortessa X-20.

FACS analysis of engineered cells overexpressing cynomolgus monkey PVRIG: The following cell lines were used to assess cross-reactivity of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) with cynomolgus monkey PVRIG (cPVRIG): parent expi and overexpressing cPVRIG expi cells. These cells were cultured in DMEM + 10% fetal calf serum + glutamax. Expi cells transiently overexpressing cPVRIG were generated by electroporation of cPVRIG DNA into parent expi cells using the Neon transfection system. For FACS analysis, expi cPVRIG cells were used within 1-3 days after transfection. The original expi cells were harvested during the logarithmic growth phase. 50,000-100,000 cells per well of each type were seeded in 96 well plates. The binding of unconjugated CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (hIgG4) and their respective controls was assessed in 8-point titration series starting at 10 μg/mL on ice for 30 min. - 1h. Titration series were performed as 3-fold dilution series. Unconjugated primary antibodies were detected using anti-human Fc antibody conjugated to Alexa 647. Data were generated using FACS Canto II or IntelliCyte and analyzed using Flow Jo and Prism software.

Cell Based Receptor-Ligand Blocking Assays: The ability of CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) to inhibit the interaction of PVRIG with its PVRL2 ligand was evaluated in a bidirectional cell competition assay.

In the first orientation, PVRL2 was endogenously expressed on non-manipulated HEK cells and the ability of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) to block binding of soluble biotinylated PVRIG Fc to HEK cells was measured. More specifically, biotinylated PVRIG Fc protein (33 nM) and CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (1.03-198 nM, hIgG4) were added sequentially to 100,000 HEK cells and incubated for 1 h on ice. The degree of binding of biotinylated PVRIG Fc was then detected by adding streptavidin Alexa 647 (Jackson Laboratories) for 20-30 min on ice. Cells were washed twice in PBS for preparation using FACS Canto II. Data was analyzed using Flow Jo, Excel (Microsoft) and Prism.

In the second orientation, HEK cells were modified to express human PVRIG (HEK hPVRIG) and the ability of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (hIgG4) to inhibit soluble human PVRL2 Fc was evaluated. More specifically, HEK hPVRIG cells were pre-incubated with CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) (0.66-66 nM) for 30 min on ice, after which PVRL2 mFc ( human PVRL2 with mouse Fc) was added (for 1 hour on ice) and its ability to bind HEK hPVRIG was measured. The extent of PVRL2 mFc binding was detected by subsequent addition of goat anti-mouse Fc A647 (Jackson Laboratories) for 20-30 min on ice. For FACS Canto II analysis, cells were washed twice in PBS. Data was analyzed using Flow Jo, Excel and Prism.

Cell Based Spatial Epitope Analysis: Epitope placement for CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) was evaluated by their ability to compete with another for binding to Jurkat cells. Briefly, Jurkat cells were harvested in logarithmic growth phase and stained (g/ml) with unlabeled CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) for 30 min on ice. Jurkat cells were then centrifuged, washed and counterstained (g/ml) with Alexa 647 labeled CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) for 30 min on ice. The competition of labeled antibodies for PVRIG binding with unlabeled antibodies on Jurkat cells was assessed by the Alexa 647 signal using flow cytometry. Data was acquired using FACS Canto II and analyzed using Flow Jo, Excel, and Prism.

results

CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) recognize PVRIG on overexpressing cells, Jurkat cells and human T cells: the ability of hybridoma humanized anti-PVRIG antibodies CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) binding to human PVRIG was evaluated with

- 67 040773 using PEC cells that overexpress human PVRIG, Jurkat cells and primary T cells. in fig. 20 shows the specificity of both CHA.7.518.1.H4(S241P) (A) and CHA.7.538.1.2.H4(S241P) (B). Both antibodies specifically bind to HEK hPVRIG cells and do not bind to original HEK cells.

Binding affinities: Both CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) also show high affinity binding to HEK hPVRIG cells with bound Kd values of 0.29 nM for CHA.7.518 .1. H4 (S241P) and 0.86 nM for CHA7.538.1.2 to bind to HEK hPVRIG cells.

In FIG. 21 shows the ability of CHA.7.518.1.H4(S241P) (A) and CHA.7.538.1.2.H4(S241P) (B) to bind Jurkat cells that endogenously express PVRIG. Both can bind Jurkat cells with comparable affinity to HEK hPVRIG cells.

The affinity of these antibodies for Jurkat cells is 0.15 nM for CHA7.518.1.H4(S241P) and 0.59 nM for CHA7.538.1.2.H4(S241P).

In FIG. 22 depicts the ability of CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) to bind CD8 T cells that have been expanded by exposure to the CMV peptide (494-503, NLVPMVATV) and which endogenously express PVRIG.

CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) detect cynomolgus monkey PVRIG (cPVRIG) expressed in expi cells: the ability of CHA7.518.1.H4(S241P) and CHA.7.538. 1.2.H4(S241P) binding to cPVRIG was evaluated using expi cells that overexpress cPVRIG. in fig. 23 shows the specificity of both CHA.7.518.1.H4(S241P) (A) and CHA.7.538.1.2.H4(S241P) (B). Both antibodies specifically bind to expi cPVRIG cells and do not bind to parent expi cells. Both CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) also show high affinity binding to expi cPVRIG cells with associated Kd values of 0.24 nM for CHA.7.518.1.H4 (S241P) and 0.58 nM for CHA7.538.1.2.

Cell Based Receptor Ligand Blocking Assays: The ability of CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) to inhibit the interaction of PVRIG with PVRL2 was evaluated in two orientations as indicated in the protocol section. In the first permutation, both CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) were able to completely inhibit PVRIG Fc binding to HEK cells (FIG. 24A). The IC50 values associated with this blocking ability are 15 nM for CHA7.518.1.H4 (S241P) and 16.1 nM for CHA7.538.1.2.H4 (S241P). It is important to note that not all hybridoma stage antibodies that have been confirmed to bind to PVRIG were able to block the binding of PVRIG Fc to HEK cells. As shown in FIG. 24B, the antibody clone designated CHA.7.544 fails to block PVRIG Fc binding to HEK cells.

In the second permutation, both CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) could also completely inhibit PVRL2 Fc binding to HEK hPVRIG cells (FIG. 25A). The IC50 values associated with this inhibition are 1.8 nM for CHA7.518.1.H4(S241P) and 2.53 nM for CHA7.538.1.2.H4(S241P). Although the ability of CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) were able to completely inhibit PVRL2 Fc binding in this permutation, consistent with their ability to inhibit PVRIG Fc binding in the first permutation, other antibodies did not show this. same trends. More specifically, another humanized hybridoma-derived antibody, CHA.7.530.3, which was able to completely inhibit PVRIG Fc binding to HEK cells (first permutation, data not shown), failed to completely inhibit PVRL2 Fc binding to HEK hPVRIG (Fig. 25A). ). Taken together, these data indicate that the second permutation in the receptor-ligand blocking assay distinguishes the activity of receptor-ligand blocking antibodies with greater sensitivity than the first permutation. Importantly, CHA.7.544 was shown to be unable to block PVRL2 Fc binding to HEK hPVRIG cells (FIG. 25B) consistent with its inability to block PVRIG Fc binding to HEK cells.

Cell Based Spatial Epitope Analysis: As indicated in the protocols section, spatial epitope analysis of CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) was performed by evaluating their ability to compete for PVRIG binding. In FIG. 26 depicts the ability of unconjugated versions of antibodies to inhibit binding to Alexa 647 (A647) conjugated versions of the same antibodies. The data in FIG. 26 shows the percentage binding of A647-conjugated antibodies relative to the maximum signal they give without competition. The signal obtained from CHA7.518.1.H4 (S241P) A647 and CHA7.538.1.2.H4 (S241P) A647 was not affected by pre-incubation of Jurkat cells with isotype control (data not shown). As expected, the signal from CHA.7.518.1.H4 (S241P) A647 and CHA.7.538.1.2.H4 (S241P) A647 was significantly reduced when competing with unconjugated versions of themselves (data not shown). Interestingly, when analyzing the A647 signal from CHA7.518.1.H4(S241P) and CHA7.538.1.2.H4(S241P) in the context of pre-incubation with the unconjugated version of the opposite antibody, a significant decrease was also observed. This indicates that CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) may have a similar epitope space on endogenously expressed PVRIG.

Summary and Conclusions: Mouse versions of anti-PVRIG antibodies designated CHA.7.518 and CHA.7.538,

- 68 040773 have been successfully humanized into the human IgG4 isotype (CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P)) which retains binding properties for the human PVRIG antigen. It has been shown that using modified overexpressing cells, Jurkat and CMV expanded primary CD8 T cells, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), are highly specific for endogenous human PVRIG and are associated with high affinity. In addition, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) also showed reactivity to the cynomolgus monkey PVRIG antigen and bound to overexpressing cells with high affinity. Functionally, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) are able to inhibit the interaction of PVRIG with PVRL2 in FACS-based assays. Finally, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) were shown to potentially share an epitope space on endogenous human PVRIG due to their ability to compete with each other for cell binding. Jurkat.

F. Example 6 Humanized Antibodies: Functional Assay of Humanized Antibodies

The functional activity of several humanized antibodies according to the invention has been confirmed.

CHO-S OKT3 assay: The CHO-S OKT3 assay was used to determine whether humanized anti-PVRIG antibodies, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), can enhance CD4+ and CD8+ T cell proliferation and secretion of cytokines. The target cells used in the co-culture assay were the Chinese hamster ovary cell line, CHO-S (ATCC), either stably overexpressing the single chain variable fragment of the anti-human CD3 antibody, clone OKT3 (abbreviated as OKT3), or stably overexpressing both OKT3 and and human PVRL2 (abbreviated as hPVRL2). Stock CHO-S OKT3 cells were grown in serum-free CD-CHO medium (Gibco) supplemented with 40 mM glutamax (Gibco), penicillin/streptomycin (Gibco), and 6 μg/ml puromycin (Gibco). CHO-S OKT3 hPVRL2 cells were grown in the same CD-CHO medium as stock cells but also supplemented with 600 μg/ml hygromycin B (Gibco).

Primary CD4+ and CD8+ T cells were isolated from healthy human donors using an enriched mixture of RosettSep™ human CD4+ T cells (Stemcell Technologies) and human CD8+ microbeads (Miltenyi Biotec), respectively, and frozen in liquid nitrogen. On the day of the CD4+ or CD8+ coculture assay, T cells were thawed, counted, and labeled with 1 μM CFSE (Life Technologies) for 10 minutes at 37°C. Following this incubation, T cells were washed and resuspended in complete medium containing RPMI (Gibco) supplemented with 10% heat-inactivated FBS, (fetal bovine serum), glutamax, penicillin/streptomycin, non-essential amino acids (Gibco), sodium pyruvate, and 50 μM β -mercaptoethanol (Gibco). CHO-S OKT3 and CHO-S OKT3 hPVRL2 cells were harvested from culture and treated with mitomycin C (Sigma-Aldrich) for 1 h at 37° C. with occasional agitation. After incubation, target cells were thoroughly washed, counted and resuspended in complete RPMI medium. The assay was performed with a 5:1 ratio of T cells (100,000) to target cells (20,000). Target cells, T cells, and 10 μg/ml of each therapy antibody were added together in a 96-well U-bottomed plate (Costar) and incubated for 3 days (CD8+ T cells) or 5 days (CD4+ T cells ) at 37°C. PVRIG antibody therapy included CHA.7.518.1.H4(S241P) human IgG4, CHA7.538.1.2.H4(S241P) human IgG4, CHA.7.530.3 human IgG4 (partial receptor/ligand blocking antibody), and CHA. 7.544 Mouse IgG1 (non-blocking receptor/ligand antibody). In addition to anti-PVRIG antibodies, the activity of mouse IgG1 against human DNAM-1 (clone DX11, BioLegend), mouse IgG1 isotype control (clone MOPC21, BioLegend) and human IgG4 isotype control were also assessed. For antibody dose titration, 3-fold dilutions of 66 nM to 0.264 nM anti-PVRIG antibodies were used and the appropriate isotype control antibody was used.

After a 3 or 5 day incubation period, co-culture supernatants were analyzed for secreted cytokines, including IL-2, IL-4, IL-5, IL-6, IL-9, IL-10, IL-13, IL-17A, IL -HF, IL21, IL-22, TNF-α and IFN-γ, with a kit for the determination of human Th1 / Th2 / Th17 cytokines (BD Biosciences) by cytometric granular analysis (CBA) or with a LEGENDplex™ Human Th cytokine kit (BioLegend ). T cell proliferation was measured by staining CD4+ or CD8+ T cells with the LIVE/DEAD Aqueous Dead Cell Staining Kit (ThermoFisher Scientific) with an anti-CD4 antibody (clone RPA-T4, BioLegend) and an anti-CD8 antibody (clone HIT8a, BioLegend ) and gated by percentage of live, CFSE-proliferating CD4+ or CD8+ T cells. Data were acquired using FACS Canto II (BD Biosciences) and analyzed using FlowJo (Treestar) and Prism (Graphpad) software.

results

CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) Enhance CD4+ T Cell Proliferation Dependent on hPVRL2: Capacity of Humanized Hybridoma Antibodies to PVRIG CHA.7.518.1.H4(S241P) and CHA.7.538 .1.2.H4 enhance the proliferation of primary CD4+ T cells in vitro was assessed using the CHO-S OKT3 assay. In FIG. 27A shows the percentage of proliferating CD4+ T cells from an exemplary donor in response to co-culture with CHO-S target cells.

- 69 040773

OKT3 hPVRL2 and various anti-PVRIG antibodies. In this donor, humanized antibodies CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P) increase CD4+ T cell proliferation compared to human IgG4 isotype control (dashed line). Partial receptor/ligand blocking antibody, human IgG4 CHA. 7.530.3 only weakly enhances T cell proliferation, while the non-receptor/ligand blocking antibody, mouse CHA.7.544 mouse IgA1, has no effect compared to isotype control antibodies. Anti-DNAM-I antibody reduces the proliferation of CD4+ T cells. in fig. 27B shows that the effects of humanized anti-PVRIG antibody CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) and anti-DNAM-1 antibody depend on overexpression of hPVRL2 on target cells. After treatment with CHA.7.518.1.H4 (S241P) and CHA.7.538.1.1 antibodies, there is a greater increase in CD4+ T cell proliferation when CD4+ T cells are co-cultured with CHO-S OKT3 hPVRL2 cells compared to co-culture with original CHO-S OKT3 cells. Similarly, anti-DNAM-1 antibody only reduces CD4+ T cell proliferation when T cells are co-cultured with hPVRL2-expressing CHO-S OKT3 cells.

CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) enhance CD8+ T cell proliferation and IFN-gamma secretion: FIG. 28A-B demonstrate the ability of the humanized anti-PVRIG antibodies CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) to increase CD8+ T cell proliferation in the CHO-S OKT3 assay. In two different donors of human CD8+ T cells, antibodies, CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.h4(S241P), increase CD8+ T cell proliferation compared to human IgG4 isotype control when T- the cells are co-cultured with CHO-S OKT3 hPVRL2 cells. However, CHA.7.544 mouse IgG1 has little or no effect. As observed with CD4+ T cells, anti-DNAM-1 antibody reduces the proliferation of CD8+ T cells. in fig. 28C shows that humanized anti-PVRIG antibodies, CHA7.518.1.H4(S241P) and CHA7.538.1.2.H4(S241P), also enhance IFN-γ secretion in CHO-S OKT3. In three different donors of human CD8+ T cells, the antibodies, CHA7.518.1.H4(S241P) and CHA7.538.1.2.H4(S241P), increase IFN-γ production compared to human IgG4 isotype control (dashed line). An increase in IL-10, IL-22 and TNF-α was also observed after treatment with CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P) (data not shown).

CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) consistently enhance CD4+ T cell proliferation in several human donors: Next, to demonstrate that CHA.7.518.1.H4(S241P) and CHA .7.538.1.2.H4(S241P) can reproducibly enhance T cell function, the effects of humanized anti-PVRIG antibodies on CD4+ T cell proliferation were examined in 11 different donors in the CHO-S OKT3 assay. In FIG. 29 shows that both CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) consistently increase CD4+ T cell proliferation in the majority of donors tested compared to the isotype control human IgG4 antibody if the T cells co-cultured with CHO-S OKT3 hPVRL2 cells. In addition, the partial receptor/ligand blocking antibody, CHA.7.530.3, and the non-receptor/ligand blocking antibody, CHA.7.544, incomparably increased CD4+ T cell proliferation in the same donors.

CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) have a dose-dependent effect on CD4+ and CD8+ T cell proliferation. Finally, the dose-dependent effect of the humanized anti-PVRIG antibodies, CHA7.518.1.H4(S241P) and CHA7.538.1.2.H4(S241P), was measured in the CHO-S OKT3 assay. Reducing the dose of antibodies CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) reduces the percentage of proliferation of CD4+ T cells (Fig. 30A) and CD8+ T cells (Fig. 30B) if the T cells coexist. cultured with CHOS OKT3 hPVRL2 cells. This dose-dependent effect on T cell proliferation is not observed with the CHA.7.544 antibody or with the IgG4 isotype control. In addition, no biphasic effect was observed with dose titration, indicating the absence of agonistic activity of the humanized anti-PVRIG antibodies.

Summary and conclusions

The humanized anti-PVRIG antibodies, CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4 (S241P), were functional in vitro in the CHO-S OKT3 cell assay. CHA7.518.1.H4 (S241P) and CHA7.538.1.2.H4 (S241P) increased CD4+ and CD8+ T cell proliferation in a dose-dependent manner. CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) also increased IFN-γ secretion in the CHO-S OKT3 assay. The activity of the antibodies, CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P), has been shown to depend on overexpression of hPVRL2 on target cells. CHA.7.518.1.H4 (S241P) and CHA.7.538.1.2.H4 (S241P) consistently increased T cell activity in several human donors, while non-blocking antibody CHA.7.544 had almost no effect.

G. Example 7 Development of Rat Monoclonal Antibodies for Mouse PVRIG

Rat monoclonal antibodies (mabs) were developed at Aldevron Freiburg (Germany). Anti-mouse PVRIG protein was increased using DNA immunization technology. An immunization vector expressing mouse PVRIG was injected into a host (rat). PVRIG

- 70,040,773 mice were expressed and an immune response developed. Positive antiserum identification and hybridoma screening were analyzed on cells that transiently express mouse PVRIG.

Creation of rat scAb against mouse PVRIG

Development of rat polyclonal antibodies against mouse PVRIG protein involved cloning the extracellular domain of mouse PVRIG into an Aldevron patented immunization vector and cloning the full length and extracellular domain into Aldevron's proprietary screening vectors. Various expression vectors used for immunization and for screening were validated by FACS in cells that transiently express mouse PVRIG. Three rats were then immunized with the immunization vector. Immune sera were taken and diluted sera were tested by FACS using cells transiently transfected with screening vectors. Blood samples from each rat were collected and purified using protein-A.

Creation of rat mAb against mouse PVRIG

Rat lymphocyte fusion and selection were performed using Aldevron test systems. This included: fusion of 20 x 96-well plates followed by initial screening with cellular ELISA (cELISA), on transiently transfected cells with mouse PVRIG (extracellular domain) or FL (full length) ECV. 108 positive clones (associated with cells expressing mouse EVA/FL PVRIG) were further expanded and retested; 30 positive clones were expanded in T-25 flasks and the supernatants were tested by cellular ELISA. 23 hybridoma clones were selected for further subcloning. Serum supernatant was tested using cELISA and FACS. A total of 21 clones were created and binding was confirmed on cells overexpressing mouse PVRIG protein.

Characteristic At

The binding of anti-mouse PVRIG antibodies of rat blood test samples, purified PCAbs, pre-cloning supernatants and clonal supernatants, as well as purified mAbs, was analyzed by flow cytometry using stable HEK293 cells overexpressing mouse PVRIG. The binding of antibodies to D10.G4.1 cells endogenously expressing mouse PVRIG was also tested. Specific expression on the cell surface of mouse PVRIG was confirmed. Cells (1-2x10 ⁵ ) were stained with viability fixing dye diluted 1:1000 in PBS for 10 min at room temperature and then the cells were washed with PBS. Abs were then added to cells (diluted in FACS buffer) and then stained with goat anti-rat-PE (diluted 1:100 in FACS buffer).

The specificity of the mAb was tested using CYRIC to transfect PVRIG cell line D10.G4.1 endogenously expressing mouse PVRIG. A decrease in cell surface was observed after PVRIG knockdown of the mouse.

Mab binding NK cells were also tested using FACS.

A sorting analysis was performed to demonstrate the diversity of mAbs.

The affinity of the purified mAbs (Kd) was determined by titrating FACS on stable cells overexpressing mouse PVRIG against cells transduced with the empty cell vector and on the D10.G4.1 cell line. Cells (1×10 ⁵ ) were stained with viability fixing dye diluted 1:1000 in PBS for 10 min at room temperature and then the cells were washed with PBS. Ab was then added to the cells (8 concentrations - 1:3 dilution series, 10-0.01 μg/ml, diluted in FACS buffer) followed by staining with goat anti-rat-PE (diluted 1:100 in FACS buffer).

Characteristic MKAT - summary table

In table. 7 (columns 1-10) shows the data obtained for the characterization of antibodies against mouse PVRIG.

Column 1 denotes ID At code

Column 2 shows the name Am provided by Aldevron

Column 3 presents the FACS data as the ratio of CIF in stably overexpressing cells to cells transduced with the empty vector at a concentration of 10 μl/ml MAb.

Column 4 shows affinity (nM) on overexpressing HEK cells

Column 5 shows binding to NK cells at a concentration of 10 µg/ml mAb

Column 6 shows the ratio of CIF binding of the D10.G4.1 cell line to the isotype control.

Column 7 shows the affinity (nM) for the D10.G4.1 cell line

Column 8 shows the different groups in the epitope-specific triage study.

Column 9 shows analysis of % receptor-ligand blocking (binding of mouse PVRIG-Fc fusion protein to mouse PVRL2-overexpressing cells) and IC50 (nM)

Column 10 shows analysis of % receptor-ligand blocking (binding of mouse PVRL2-Fc fusion protein to mouse PVRIG-overexpressing cells) and IC50 (nM)

- 71 040773

AB-406 and 0AB-47, which have shown blocking activity in receptor-ligand binding assays, have relatively high affinity to bind to NK and D10.G4.1 cells.

These Abs were chosen for TME expression and for in vivo studies.

Table 7. Characterization of mouse anti-PVRIG monoclonal Ab

LI MS ID Name At CIF Ratio (NEC OX/EV) 10 µg/mL Kd (nM) on OH cells Expression in NK, MFIr (Ab/iso) Express to D10.G4.1 CIF ratio (at/iso) Kd (nM) at D10. G4.1 cell ah epitope-specific sorting % R-L blocking (029-Fc) (IC50) % R-L Blocking NIA (mPVRL2Fc) (IC50) AB 400 BOJ- 1F11- H6 10 0.139 3 BUT 13.2 35.68 1 Agonist Agonist AB 401 BOJ- 3E2F4 55.7 2.4 BUT 7.5 BUT 1 Agonist Incompatible AB 402 BOJ- 4F11- H6 4.8 0.089 74 BUT 20 BUT 1 Agonist Agonist AB 403 BOJ- 4G1- E3 28 57.64 - 2 BUT 4 78% (2.782) Incompatible AB 404 BOJ- 4H8EZ 18.6 0.386 833 1.5 6.8 BUT 1 Inert 93-98% AB 405 BOJ- 5A4EZ 13.5 0.088 71 BUT 6.7 4.596 1 Agonist Agonist AB 406 BOJ5С7vs 54.5 1.884 667 2.2 10 8.577 3 77% (3.679) 97-100% AB 407 BOJ- 5G4F4 50 0.334 427 2.8 8 1.325 3 95% (3.992) 95-100% AB 408 BOJ- 8G1- G1 18.9 5.098 1.7 8 BUT 4 90% (3.585) Agonist AB 409 BOJ- 9B1- D9 24.8 0.155 5 BUT 10 BUT 1 Agonist Incompatible AB 410 BOJ11С2G9 16.6 0.221 8 2.4 6.8 BUT 1 Agonist Incompatible AB 411 BOJ- 12E2F8 36 5.340 5 BUT 5.4 BUT 3 94% (4.311) 96-100% (1.189) AB 412 * BOJ14H2F4 3.3* BUT BUT 2 BUT 2 Agonist 62-67% (23.87) AB 413 * BOJ- 15VZ- E11 49.3 1.093 45 2.1 12 4.693 3 73% (2.836) 95-100% (0.58) AB BOJ- 24.8 2.039 1.6 8.5 16.48 3 90% 95-100%

414 15F8- C6 5 (2.907) (1.164) AB 415 BOJ- 16E7G8 24 17.57 3 2.2 5 BUT 5 95% (6.727) 93-100% (5.643) AB 416 BOJ17С4D4 26.3 6.357 1.5 5 BUT 1 Inert 92-94% (1.963) AB 417 * BOJ17С7- H5 15 59.16 - 2.5 BUT 4 72% (3.455) Inert AB 418 BOJ18C1SU 64.8 62.98 BUT 4.4 BUT 5 94% (18.18) 100% (8.055) AB 419 BOJ18D2- F5 17 0.316 9 1.5 6.3 BUT 1 Agonist 57-64% AB 420 BOJ- 19D9С7 33.6 3.172 1.2 20 BUT 4 78% (9.805) Agonist

- 72 040773

H. Example 8 Combination Testing with Additional Immune Checkpoint Inhibitors

State of the art

Although blockade by antibodies to the CTLA4 and PD1 pathways has become an effective cancer treatment, most patients do not receive a long-term therapeutic effect, indicating the need to target additional immune system checkpoints. Using our unique computational algorithms to identify new members of the B7/CD28 family, we have identified PVRIG, which is expressed by several subpopulations of T and NK cells. In this document, we report its expression profile, functional characterization, and antitumor activity of blocking antibodies targeted to this molecule.

Ways

Using the Predictive Discovery platform, PVRIG was identified as a potential new immune checkpoint, after which a retroviral cell screening library was used to identify the cognate binding partner. Target effects on T cell modulation were assessed by primary and tumor-derived T cell assays, taking advantage of target overexpression, knockdown, and antibody antagonist approaches. Anti-human protein antibodies were screened for their ability to increase T cell activation in vitro, while antibodies targeting the mouse orthologue were evaluated in vivo for tumor growth inhibitory effects in syngeneic models.

results

The PVRIG-Fc fusion protein was found to bind PVRL2 with binding specificity confirmed by both ELISA and flow cytometry analysis. PVRIG has demonstrated unique expression kinetics upon T cell activation, with targeting on memory T cells as well as NK cells and γδ T cells. A group of high affinity human antibodies with the ability to block the interaction of PVRIG with PVRL2 has been shown to show increased activation of both primary CD4+ and tumor CD8+ T cells via a PVRL2-dependent mechanism when tested in vitro.

Because CHA.7.518.1.H4 (S241P) is not cross-reactive with mice, in vivo studies were performed with a surrogate blocking antibody against mouse PVRIG. In combination with anti-PDL-1 blockade, mouse anti-PVRIG inhibits the growth of established tumors in both CT26 and MC38 rectal cancer models. Combination testing with additional immune checkpoint inhibitors, as well as in PVRIG knockout mice, is ongoing.

conclusions

A high affinity antagonist antibody is able to enhance human T cell activation, and a surrogate antibody with similar characteristics shows synergy with PD-L1 in vivo in several syngeneic models. Overall, our data demonstrate the feasibility of targeting PVRIG in addition to other B7 family checkpoints for cancer treatment.

I. Example 9: In vivo experimental proof of concept: anti-MPVRIG mAb efficacy in the CT26 tumor model

This example describes the efficacy of anti-mPVRIg mAb treatment in a mouse model of CT26 colon carcinoma as monotherapy or in combination with anti-PDLL therapy.

Materials and methods

Tumor Infection Experiments:

CT26 colon carcinoma was purchased from ATCC (CRL-2638). Cells were cultured in RPMI 1640 (Biological Industries, 01-100-1 A) with 10% FBS (Biological Industries, 04-127-1A) and 100 μg/ml penicillin/streptomycin (Biological Industries, 03-031-1B). For tumor implantation, cells were harvested, washed, counted and suspended to 10 ⁷ cells/ml in cold RPMI 1640 and placed on ice. BALB/c mice ((female, 8 weeks old) Envigo) were anesthetized with 10% ketamine (Clorketam, SAGARPA Q-7090-053) and 10% xylazine (Sedaxylan; BE-V254834) injected intraperitoneally. The backs of the mice were then shaved and disinfected with a 70% ethanol solution. Tumor cells were injected as 50 μl of 5 x 105 CT26 cells subcutaneously into the posterior right flank of mice. The introduction of mAb was started on the 4th day (monotherapy) or the 7th day (combination therapy) after tumor inoculation, when the tumors had a volume of 30-50 mm ³ (monotherapy) or reached a volume of 60-90 mm ³ (combination therapy); and administered intraperitoneally (i/p) in a final volume/injection of 200 μl, for 3 weeks for a total of 6 injections. Tumor growth was measured with an electronic caliper every 2-3 days and described as 0.5 x W ² x L mm ³ . Mice were sacrificed with CO ₂ at any end of the study or at any of the following clinical endpoints: tumor volume >2250 mm ³ , tumor ulceration, body weight loss > 20%, or dying appearance

Antibodies:

It has been shown that chimeric antibodies against mouse PVRIg (mAb 406 and mAb 407) used in

- 73 040773 in this study, designed as an isotype rat IgG2b monoclonal antibody (mAb), binds to mPVRIg-expressing 293HEK transfectants and blocks mPVRL2 binding to these cells. The anti-murine PDL-1 mIgG1 inhibitor used in this study was mAb YW243.55.S70. Antibody YW243.55.S70 is an αanti-PD-L1 described in WO 2010/077634 (heavy and light chain variable region sequences depicted in SEQ ID NOs: 20 and 21, respectively, of WO 2010/077634) and has the sequence, described in it.

All mAbs were formulated in sterile PBS with low endotoxin content (<0.05 EU/mg). Table 8. Tested mAb.

Mouse IgGl, to isotype BP0083 BioXcell control. (MORS-21)

Rat IgG2b, to isotype BP0090 BioXcell control.

(LTF-2,) Compugen Inc. Reference anti-PDL-l YW243.55 (mlgGl) .S70 Anti-CGEN PVRIG mAb 406 BOI-5C7- ALDEVERON (rat IgG2b) VZ Anti-CGEN PVRIG mAb 407 BOI-5G4- ALDEVERON (rat IgG2b) F4

Study Design Monotherapy

Eight-week-old female BALB/c mice were purchased from Envigo and kept in a pathogen-free (SPF) vivarium for 1 week prior to the start of the experiment. Mice were anesthetized, shaved and inoculated subcutaneously with 50 μl of 5×10 ⁵ CT26 tumor cells. On day 4 post-tumor inoculation, mice were randomized into treatment groups of n=10 (as described below). Mice received mAb treatment (as detailed below) by injection on days 4, 7, 11, 14, 18 and 21 post-inoculation. Tumor growth was measured with a caliper every 2-3 days.

Table 9. Treatment groups

No. groups Treatment/mAb Dose (mg/kg) dose number Volume / dose (µl) 1 Carrier 6 200 2 mlgGl iso Ctrl 5 6 200 3 Rat IgG2b iso Ctrl 10 6 200 4 Anti-PDL-l mlgGl 5 6 200 5 Anti-mPVRIg mAb 406 rIgG2b 10 6 200 6 Anti-mPVRIg mAb 407 rIgG2b 10 6 200

Combination Therapy

For combination therapy of anti-mPVRIg and anti-mPDL-l mAbs. Mice were treated as described in monotherapy. On day 7 post-tumor inoculation, mice were randomized into treatment groups of n=10 as described below. Mice received mAb treatment (as detailed below) by injection on days 7, 11, 14, 18, 21 and 25 after tumor inoculation.

Table 10. Therapeutic doses

# Group Treatment/mAb 1 Dose (mg/kg) Treatment/mAb 2 Dose (mg/kg) # Dose Volume/up per (µl) 7 mlgGl iso Ctrl 5 Rat IgG2b iso Ctrl 10 6 200 8 Anti-PDL-l mlgGl 5 Rat IgG2b iso Ctrl 10 6 200 9 Anti-PDL-l mlgGl 5 Anti-mPVRIg mAb 406 rIgG2b 10 6 200 10 Anti-PDL-l mlgGl 5 Anti-mPVRIg mAb 407 rIgG2b 10 6 200

Statistical analysis:

Two-way repeated measures ANOVA followed by two-way repeated measures ANOVA for selected pairs of groups using JUMP (Statistical DiscoveriesTM) software. Analysis of tumor growth measurements was performed by comparing tumor volumes measured on the last day when all study animals were still alive. Statistical differences in percent from mice without tumors were determined using

- 74 040773 using the Mantel-Cox logarithmic rank test. P values <0.05 were considered significant.

* p<0.05; ** p<0.01; *** p<0.001. For each experiment, the number of repetitions performed and the number of animals per group are described in the corresponding legend(s) of the figure(s) (FIGS. 47-48).

results

Monotherapeutic activity of anti-mPVRIg and anti-mPDL-l in syngeneic CT26 tumor model

We have begun preclinical evaluation of anti-mPVRIg and anti-mPDL-l monotherapy in the mouse syngeneic CT26 tumor model. We treated mice with anti-PDL-1 antibody of mIgG1 isotype (YW243.55.S70) or with anti-mPVRIg of rIgG2b isotype (mAbs 406 and 407).

In a semi-therapeutic treatment model for CT26 colon carcinoma, anti-PDL-1 monotherapy was significantly effective (P<0.0001), causing 70% TGI compared with mIgG1 isotype control, higher tumor rejection rates with rapid tumor rejection and prolonged antitumor efficacy. immunity observed in most mice (Fig. 63A+B).

The anti-mPVRIg mAb 406 and anti-mPVRIg 407 mAb treated groups showed similar tumor growth rates without TGI compared with the rIgG2b isotype (Fig. 63A+B). Accordingly, treatment with anti-PDL-1 mIgG1 prolonged the survival of mice (P<0.01, Fig. 63c), with 5 out of 10 mice showing complete tumor clearance (Fig. 63b). No effect of anti-mPVRIg mAbs on survival was found.

Activity of a combination of anti-PVRIg and anti-PDL-1 in a syngeneic mouse tumor model

We then evaluated the activity of anti-PVRIg and anti-PDL-1 combination therapy in a syngeneic mouse tumor model.

In a model of therapeutic treatment of CT26 colon carcinoma, the administration of anti-PDL-1 with rIgG2b isotype control treatment, initiated on the 7th day after inoculation, was not effective, while the combination of anti-PVRIg mAb 407 with anti-PDL-1 caused a significant TGI. (56%, P=0.0005), higher rates of tumor rejection in 4 out of 10 individuals showing complete tumor clearance (Fig. 64A+B) and contributed to better antitumor activity with sustained antitumor immunity detected (P<0.01, Fig. 64C). The combination of anti-PVRIg mAb 406 with anti-PDL-1 was partially effective, resulting in a 33% TGI, but the reported antitumor response was transient and did not affect survival.

conclusions

mPVRIg was expected to play the role of a novel B7-like molecule and thus a potential target for antibody-based immunotherapy. Several human in vitro experimental systems have demonstrated an immunomodulatory effect for mPVRIg. In the studies presented in this report, we evaluated the in vivo anticancer effect of mAb directed against mPVRIg. In our study, treatment with 10 mg/kg (200 μg/mouse) of anti-mPVRIg as monotherapy with minimal disease progression, i.e., initiation of treatment on day 4 (mean tumor 40 mm ³ ), did not result in TGI or survival at that time. while the anti-PDL-1 mAb positive control showed significant TGI and resulted in prolonged survival.

The anti-mPVRIg mAb was also tested in combination with αHntu-PDL-1 treatment. Treatment with 10 mg/kg (200 μg/mouse) was started on day 7, when the tumors reached an average size of 75 mm ³ . Combination therapy of anti-mPVRIg mAb 407 with anti-PDL-1 in the CT26 therapeutic model demonstrated tumor growth inhibition and prolonged survival in treated mice. The effect on tumor growth varied between individual mice: some individuals showed complete tumor clearance, while others showed a partial response (temporary TGI), and some individuals did not respond. The in vivo effect of anti-mPVRIg and anti-mPDL-1 combination treatment was also shown in the MC38 and B16-D ^b /gp100 tumor models (data not shown).

Additional in vivo studies are planned to evaluate dose-response and efficacy in additional syngeneic models or in combination with additional therapeutic compounds or treatment regimens.

J. Example 10: Discovery of a Therapeutic Antibody to TIGIT Using Phage Display

1. Introduction

A phage display antibody detection campaign was conducted to isolate human TIGIT binders from a naive fab library using the human TIGIT recombinant extracellular domain as the target antigen. Forty-five new antibodies specific for human TIGIT have been isolated and generated as human IgG4, including the optional S241P in the hinge region, as discussed herein. The resulting antibodies were screened for their ability to block the TIGIT-PVR interaction and for cross-reactivity with cynomolgus TIGIT expressed by flow cytometry. Two of these antibodies were further optimized to increase the binding affinity of human and cynomolgus TIGIT.

- 75 040773

2. Protocols

Antigens for Antibody Detection by Phage Display: Two human TIGIT protein formats were used as antigens in phage display. The first of these consisted of a human TIGIT EVA (Met22 - Pro141) fused to a C-terminal polyhistidine tag (hTIGIT-HIS) and was either self-produced or commercially obtained from Sino Biological Inc. The second antigen format, consisting of a human TIGIT EVA fused to a human IgG1 Fc domain at the C-terminus (hTIGIT-hFc), was either produced in-house or obtained commercially from R&D Systems.

Functional OC of antigens: The recombinant TIGIT antigens used for biopanning have been functionally validated for their ability to bind to human PVR, a human TIGIT ligand. Biotinylated antigens were tested for PVR binding either by ELISA or by flow cytometry. Biotinylated hTIGIT-HIS was confirmed for its ability to bind hPVR-hFc (Sino Biological Inc.) by ELISA. Biotinylated hTIGIT-hFc was confirmed by flow cytometry for its ability to bind endogenously surface-expressed PVR on Expi293 cells.

Phage panning of a human antibody library: Two phage campaigns were performed using either human TIGIT-HIS (campaign 1) or human TIGIT-hFc (campaign 2) as antigens. Panning reactions were performed in solution using streptavidin-coated magnetic beads to display biotinylated TIGIT antigens. Both campaigns used a human fab antibody phage display library for initial detection. Three rounds of panning were performed using the corresponding human TIGIT antigens with higher wash and lower antigen concentrations for each subsequent round of panning. The CPA.9.002 antibody generated in Campaign 1 was optimized to improve human TIGIT binding by generating a phage library by LCDR3 saturation mutagenesis and panning the resulting library against human TIGIT-HIS (Campaign 3). Two antibodies, CPA.9.059 and CPA.9.027, generated in campaigns 2 and 3, respectively, were also optimized to improve affinity for human TIGIT and cynomolgus TIGIT cross-reactivity (campaign 4). For each antibody, a phage library was generated by saturation mutagenesis of two CDRs (any combination of H-CDR1, H-CDR2, H-CDR3, L-CDR1, or L-CDR3). The resulting phage libraries were subjected to four rounds against human TIGIT-HIS and C-terminal HIS-labeled cynomolgus monkey TIGIT recombinant protein in alternating rounds of panning. The antigens used for panning were as follows: 1 nM human TIGIT-HIS in round 1, 1 nM cynomolgus TIGIT-HIS in round 2, 0.1 nM human TIGIT-HIS in round 3, and 0.1 nM cynomolgus TIGIT-HIS in round 3. round 4.

Binding screen using antibodies expressed as fab fragments: The phagemid construct contains an amber stop codon that allows it to function as a fab expression vector. Transformation of these vectors into E. coli and induction with isopropyl-D-I-thiogalactopyranoside (IPTG) results in periplasmic expression of soluble Fab molecules. Fab proteins secreted into the E. coli periplasm were extracted by osmotic shock for binding screening.

Primary screening by ELISA: PPE fab extracts were tested for binding to the hTIGIT-HIS or hTIGIT-hFc panning antigen by ELISA. Positive results from the ELISA screen were sequenced using heavy chain and light chain specific primers. The sequences were collected and analyzed. Clones were considered unique in sequence if several non-conservative differences in heavy chain CDR3 were noted.

Secondary screening by flow cytometry: sequence-unique ELISA-positive fab clones were selected and analyzed for their ability to bind Expi293 cells overexpressing human TIGIT by flow cytometry. Stock Expi293 cells were used as a negative control for each fab sample.

Positive fab reformatting and production as human IgG4 molecules: Potential human TIGITs that bind fab were converted to full-length human IgG4 (including the S241P hinge mutant, see Aalberse et al, Immunology 202 105: 9-19, incorporated herein by reference in general, and in particular for the discussion of S241P, and references 1, 2 and 3 cited therein) for further characterization. The protein expression constructs were generated by PCR amplification of the heavy and light chain variable region sequences, which were subcloned into the pUNO3 vector (Invivogen).

3. Results

Functional OS of Recombinant Human TIGIT Proteins: The recombinant hTIGIT-HIS and hTIGIT-hFc proteins, either self-produced or commercially produced, have been functionally validated for their ability to bind to human PVR. Human PVR (Fc-conjugated) showed dose-dependent binding to biotinylated hTIGIT-HIS in ELISA (data not shown). Similar binding was observed in reverse orientation where PVR was immobilized on an ELISA plate and hTIGIT-HIS was in solution (data not shown).

- 76 040773

The hTIGIT-hFc protein was functionally confirmed by binding to PVR in a flow cytometry assay. In this assay, the hTIGIT-hFc protein was titrated against Expi293 cells that endogenously express human PVR. The interaction was detected using an anti-hFc secondary antibody conjugated to the AF647 fluorescent label. An irrelevant Fc protein was used as a control (data not shown).

Functional analyzes were performed on a number of candidates as described in the examples below.

Affinity maturation binding screens using antibodies expressed as fab fragments: Eight 96-well plates with periplasmic-extracted clones were analyzed for de novo campaigns (1 and 2). Seventy-three unique clones were identified in Campaign 1 using the hTIGIT-HIS protein as the target antigen. A secondary screening of 73 ELISA-positive clones by flow cytometry revealed 21 positive results for binding to Expi293 cells overexpressing TIGIT. A similar screen for Campaign 2 (hTIGIT-HFC as target antigen) produced 37 ELISA positive clones, 24 of which were also positive for binding to human TIGIT overexpressing Expi293 cells by flow cytometry (FIG. 52).

Two 96-well plates of fab clones (as PPE) were screened for optimization/affinity campaigns (3 and 4). ELISA-positive unique variants were screened for binding to Expi293 cells overexpressing human and/or cynomolgus TIGIT by flow cytometry. The binding affinities of the upstream clones to the hTIGIT-HIS protein were also assessed by surface plasmon resonance (SPR). The first cycle of affinity maturation of the CPA.9.002 antibody produced 5 new antibodies, CPA.9.021, CPA.9.027, CPA.9.044, CPA.9.048 and CPA.9.049, with mutations in L-CDR3 and at least a 3-fold improvement in binding affinity to recombinant human TIGIT. The second round of optimization of the CPA.9.027 antibody yielded 4 new antibodies with at least a 25-fold improvement in binding to recombinant human TIGIT. The new variants showed mutations in H-CDR2 and L-CDR3 (CPA.9.083 and CPA.9.086) and additionally in L-FR4 for CPA.9.089 and CPA.9.093. Optimization of CPA.9.059 resulted in two new antibodies, CPA.9.101 and CPA.9.103, with significantly improved binding to cynomolgus TIGIT as well as significant improvement in human TIGIT binding for CPA.9.103. Mutations were observed in H-CDR3 and L-CDR1 for both new variants. In addition, minor changes in L-FR1 were observed for CPA.9.101.

Reformatting of ELISA- and FACS-positive fab into hIgG4: Forty-five unique fabs positive for binding by ELISA and flow cytometry to human TIGIT were reformatted for expression as human IgG4 molecules, including the optional S241P hinge variant, as discussed herein. In addition, 11 affinity optimized variants were also reformatted as IgG4. The sequences of selected phage-derived antibodies are shown in FIG. 53. The sequences of the two reference antibodies, BM26 (WO 2016/028656, clone 31C6) and BM29 (US 2016/0176963, clone 22G2) are also shown in FIG. 53 for comparison. The reformatted antibodies were evaluated for binding to Expi293 cells overexpressing human TIGIT and a binding curve was created to calculate the equilibrium binding constant (KD). These antibodies were also evaluated for binding to Expi293 cells overexpressing cynomolgus monkey TIGIT, as well as their ability to block the interaction between human TIGIT and human PVR in cellular assays. Based on these characteristics, a subset of these antibodies were selected for in vitro functional assays, as described in more detail below.

K. Example 11: Discovery of a Therapeutic TIGIT Antibody Using Hybridoma

1. Rationale and objectives

Hybridoma technology, using known and standard methods in the art, has been used to generate mouse antibodies that bind to human TIGIT with high affinity, cross-react with non-human primate TIGIT (cynomolgus macaque, Masaca fascicularis, called supo), and block the interaction. TIGIT with its ligand, PVR (CD155).

2. The essence of the invention

Balb/c mice were immunized with recombinant forms of human and cynomolgus TIGIT extracellular domain proteins. Cells isolated from the spleen and lymph nodes of immunized mice were fused with an Sp2/0 myeloma cell line to create hybridomas that secrete mouse antibodies. Supernatants from polyclonal and subcloned monoclonal hybridomas were screened for binding to Expi293 cells overexpressing human and cynomolgus TIGIT and for binding affinity to recombinant human and cynomolgus TIGIT proteins using standard SPR methods. Mouse antibodies from selected hybridomas were purified and characterized in detail in binding assays and functional assays. Five functional and cynomolgus cross-reactive mouse antibodies were humanized to contain the hIgG4 framework (including the optional hinge variant as described herein) and isotype. The sequences are shown in Fig. 53.

L. Example 12: FACS Binding KD Determinations of Phage and Hybridoma Derived Antibodies Associated with Cells Overexpressing Human and Cynomolgus TIGIT

- 77 040773

1. Protocols

The following cell lines were generated to evaluate the binding affinity of human phage and mouse anti-TIGIT antibodies: Expi293 native, Expi293 overexpressing human TIGIT, and Expi293 overexpressing cynomolgus TIGIT. The following hybridoma and phage antibodies were prepared in 11-point 2-fold dilution series over a concentration range at the binding site of 195 pM-200 nM:

Phage derived antibodies: CPA.9.027, CPA.9.049, CPA.9.059.

Hybridoma derived antibodies (before humanization): CHA.9.536, CHA.9.541, CHA.9.543, CHA.9.546, CHA.9.547 and CHA.9.560. Two different reference antibodies were included: BM26 (W02016/028656A1, clone 31C6 as mouse IgG1) and BM29 (US2016/0176963 A1, clone 22G2 as mouse IgG1).

The 12th well of each titration contained a buffer to serve as background. Each cell type was incubated with anti-human TIGIT mAb for 60 min at 4°C. After washing, AF647-labeled goat anti-human F(ab') (Jackson Immunoresearch) and AF647-labeled goat anti-mouse IgG-Fc (Southern Biotech # 103030), respectively, were added to cells incubated with human and mouse mAbs, respectively. After that, the geometric mean fluorescence intensity (gSIF) of 5000-10000 events for each well was recorded on a FACS Canto II HTS instrument. A plot of gSIF as a function of molecular concentration of human PVR was aligned using a single site model, Graphpad Prism specific binding to estimate KD, and 95% confidence intervals of each non-linear fit.

2. Results

Two independent KDs by the FACS method, measured for each mAb, differed on average by no more than 2 times. One illustrative measurement for KD, together with a 95% confidence interval approximation of the binding isotherm, recalculated for each mAb for overexpressing human and cynomolgus monkey cells in FIG. 54 and 55 respectively. CPA.9.059 showed no binding to overexpressing cynomolgus monkey cells. It should be noted that binding site concentrations (2X molecular concentration) for all mAbs are used for non-linear fitting of the curve, which means that the assumption made is that this FACS KD method measures the binding site constant (kD) and not the molecular or stoichiometric binding constant.

M. Example 13: FACS blocking assay of phage and hybridoma derived anti-human TIGIT mAbs that inhibit PVR-FC binding to TIGIT 1.

Introduction

The purpose of this assay is to characterize the ability of phage and hybridoma-derived anti-human TIGIT antibodies to inhibit the binding of human PVR to human TIGIT overexpressed on the cell surface. First, the binding affinity for human PVR human TIGIT will be determined using FACS. Binding isotherms showed the saturation concentration of human PVR, which was used for blocking assays. Cells overexpressing human TIGIT were then titrated with anti-TIGIT mAbs produced by phage and hybridoma, followed by the addition of a saturating concentration of human PVR. The binding of anti-human TIGIT antibody to overexpressing cells was then measured using FACS.

2. Protocols

FACS KD analysis.

Various isotypes of human PVR-Fc were tested by FACS for optimal binding and human PVR-h1Fc (Sino Biological # 10109-H20H) was determined, and human PVR-m2aFc (Compugen) showed the highest levels of binding to cells overexpressing human TIGIT. The two PVR isotypes were each diluted by 2-fold serial dilution over 11 point titrations to a final molecular concentration range of 98 pM-100 nM. The 12th well of each titration contained only buffer to serve as background. Each cell type was incubated with mAbs for 60 min at 4°C. After washing, AF647-labeled F(ab')2 goat anti-human Fc fragment (Jackson Immunoresearch # 109-606-098) and AF647-labeled goat anti-mouse IgG (SouthernBiotech # 1033-31) were added to wells that were titrated with anti -TIGIT mAb human and mouse, respectively. After that, the geometric mean fluorescence intensity (gSIF) of 5000-10000 events for each well was recorded on a FACS Canto II HTS instrument. A plot of gSIF as a function of molecular concentration of human PVR was aligned using a single site model, Graphpad Prism specific binding to estimate KD, and 95% confidence intervals of each non-linear fit. The results of human PVR-m2aFc and human PVRh1Fc are shown in FIG. 57A and B, respectively.

Phage mAb blocking assay: The following hIgG4 phage antibodies and reference mAbs were prepared in three-point 5-fold dilution series in the binding site concentration range of 267 pM - 6.7 nM: CPA.9.027, CPA.9.049 and CPA.9.059, and BM26 ( WO2016/028656A1, clone 31C6 as hIgG4) and synagis hIgG4 (negative isotype control).

- 78 040773

The 4th well of each titration contained only buffer to serve as background. Cells were incubated with mAb for 15 min at 4°C. Human PVR-m2aFc (Compugen) was then incubated for 1 h at 4°C. After washing, AF647-labeled goat anti-mouse IgG (SouthernBiotech # 1033-31) was added. After that, the geometric mean fluorescence intensity (gSIF) of 5000-10000 events for each well was recorded on a FACS Canto II HTS instrument. The hSIF values of human bound PVR for cells pre-incubated with mAb were compared with the hSIF values of cells pre-incubated with blocking reference mAb and non-blocking control mAb. If a phage antibody reduced the human PVR-m2aFc binding signal compared to the signal from a titration with a known non-blocking MAb, then the antibody was characterized as blocking PVR binding at that concentration of phage MAb. Patterns of blocking phage mAbs were similar to blocking PVR reference BM26 (Fig. 58).

Hybridoma mAb blocking assay: The following hybridoma antibodies were prepared in 11-point 2.5-fold dilution series over a binding site concentration range of 14 pM-133 nM: CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547, CHA. 9.560, BM26 (WO 2016/028656A1, clone 31C6 as mouse IgG1) and BM29 (US2016/0176963 A1, clone 22G2 as mouse IgG1).

The 12th well of each titration contained only buffer to serve as background. Cells were incubated with mAbs for 15 min at 4°C. Human PVR-h1Fc (Sino Biological # 10109-H20H) was then added and the cells were then incubated for 1 hour at 4°C. After washing, an AF647-labeled F(ab') ₂ fragment of a goat anti-human Fc antibody (Jackson Immunoresearch) was added. After that, the geometric mean fluorescence intensity (gSIF) of 5000-10000 events for each well was recorded on a FACS Canto II HTS instrument. A plot of hSIF as a function of mAb binding site concentration was non-linearly fitted using Graphpad Prism's log(inhibitor) vs. slope-response (four-parameter) model to estimate the IC50 of each non-linear fit. This experiment was repeated twice over two days.

3. Results

In FIG. 58 and 59 show that both phage and hybridoma antibodies strongly block the binding of human PVR-Fc to human TIGIT overexpressed on the cell surface of Expi293 cells. The blocking activity of the phage and hybridoma antibodies is comparable to the two reference antibodies tested, BM26 and BM29.

N. Example 14: Surface Plasmon Resonance (SPR) Kinetic Studies of the Binding of Nine Phage and Hybridoma-Derived MABs to Human, Cynomolgus, Mouse TIGIT

1. Protocols

All experiments were performed using a ProteOn XPR 36 instrument at 22°C. First, high-density capture surfaces were generated with goat anti-human Fc polyclonal antibody (Thermo #HI0500) and rabbit anti-mouse antibody (GE Healthcare #BR100838), respectively, immobilized across all vertical capture bands and horizontal intermediate points on separate GLC chips. using a standard amine bond. Typical immobilization levels for anti-human capture PCAt and anti-mouse capture antibody for each GLC chip were about 5000RU. The human TIGIT was obtained from Sino Biologicals, while the mouse TIGIT monomer and the cynomolgus TIGIT monomer were independently obtained. Purified mAbs tested for binding to human, mouse and cynomolgus TIGIT are listed below:

Phage antibodies: CPA.9.027, CPA.9.049 and CPA.9.059

Hybridoma antibodies: CHA.9.536, CHA.9.541, CHA.9.543, CHA.9.546, CHA.9.547 and CHA.9.560

Comparative tests: BM26 (W02016/028656A1, clone 31C6 as hIgG4) and BM29 (US2016/0176963 A1, clone 22G2 as hIgG4).

Each mAb was diluted to ~0.5 µg/mL in running buffer, which was lxPBST with filtered BSA added to a final concentration of 100 µg/mL. For each cycle of single kinetics on the ProteOn instrument, another mAb was captured in one of the six unique vertical capture strips for approximately 1.5-2.5 minutes. After switching the ProteOn buffer flow to the horizontal direction, the capture surfaces stabilized for about 15-20 minutes. Six concentrations of 3-fold dilution series of human TIGIT (346 pM - 84.1 nM), cynomolgus TIGIT (371 pM - 90.2 nM), or mouse TIGIT (382 pM - 92.9 nM) were administered over 2 min, and followed by 20 min dissociation at a flow rate of 50 µl/min. An identical injection buffer preceded each series of injected antigen for double comparison. Anti-human antibody surfaces were regenerated with two 30 second injections of 146 mM phosphoric acid, and anti-mouse antibody capture surfaces were regenerated with two 30 second pulses of 10 mM glycine, pH 1.7. Sensorgrams of the TIGIT antigen administered over the captured mAbs were processed using the Scrubber ProteOn version and fit a 1:1 kinetic binding model, including the term for mass transfer.

In FIG. 56 shows the obtained kinetic rate constants and equilibrium constants

- 79 040773 dissociation, where the data were sufficiently reliable to estimate the binding constants (sensogram data not shown). The asterisks indicate kd values, which must be maintained at a constant level of 1.0x10-5/s. In cases such as binding of clone CHA.9.560 to human TIGIT, the kinetic model was able to estimate Kd, but it is practically impossible to accurately estimate Kd of the order of 1 x 10 ^-6 /s after only 20 min of dissociation data, given the sensitivity of the instrument.

A. Example 15: Anti-TIGIT Antibody Functional Assays

1. Rationale and objectives

In order to functionally characterize the ability of human anti-TIGIT antibodies to inhibit the interaction of TIGIT and its PVR ligand and therefore enhance human T-cell activation both alone and in combination with human anti-PVRIG antibody, CHA.7.518. 1.H4 (S241P).

2. Protocols

Human TIGIT/CD 155 Jurkat IL-2 Luciferase Reporter Gene Assay: The Jurkat IL-2 Human TIGIT Luciferase Reporter/PVR Bioassay Kit (Promega) was used to evaluate the effect of anti-human TIGIT antibody treatment on T cell activation. Jurkat T cells were stably transfected with recombinant human TIGIT and the luciferase reporter gene driven by the IL-2 response element (IL-2-RE). The stimulating cells were artificial CHOK1 APC cells (aAPC) expressing recombinant human PVR and an engineered cell surface protein designed to activate TCR-mediated signaling in an antigen-independent manner. After co-culture of these cells, the human TIGIT/human PVR interaction inhibits TCR signaling and IL-2-RE mediated luminescence. The addition of an anti-human TIGIT antibody that blocks the human TIGIT/human PVR interaction releases an inhibitory signal, resulting in T cell activation and IL-2-RE-mediated luminescence. The assay was performed according to the manufacturer's instructions. Briefly, human Cho-Κι PVR aARC cells were thawed in a water bath at 37° C. and diluted in F-12 medium supplemented with 10% FBS (Promega). 25,000 cells/well were plated on white, flat bottom prepared 96 well tissue culture plates (Costar). The plates were then incubated overnight at 37°C. The following day, hybridoma- and phage-derived anti-human TIGIT antibodies, mouse IgG1 (mIgG1), and hIgG4 isotype control antibodies or reference (BM) anti-human TIGIT antibodies were added either as a single dose at 10 μg/mL or in 10-point, 2-fold dilution series starting at 20 µg/ml. Jurkat IL-2-RE human TIGIT luciferase cells were thawed in a water bath at 37° C. and diluted with RPMI medium supplemented with 10% FBS (Promega). 125,000 Jurkat cells were added to each well. The plates were then incubated at 37° C. with 5% CO ₂ for 6 hours. After incubation, the plates were removed from the incubator and left to equilibrate to room temperature for 30 minutes. To each well was added 80 μl of Bio-Glo luciferase substrate (Promega) and the mixture was left to equilibrate for 10 min at room temperature protected from light. Luminescence was quantified on an EnVision multichannel reader (Perkin Elmer) with an ultrasensitive luminescent detector. The luminescent signal was presented in relative light units (RLU).

Propagation of CMV-specific human CP8 ⁺ T cells: Human CMV-reactive peripheral blood mononuclear cells (PBMC) (CTL) were thawed, resuspended at 2x106 cells/ml and stimulated with 1 μg/ml CMV pp65 peptide (Anaspec) in complete RPMI medium with adding 2 ng/ml recombinant human IL-2 (R&D systems) and 10 ng/ml recombinant human IL-7 (R&D systems) at 37°C. After 9 days, the cells were separated 1:2 and the cell cycle was stopped with a low dose of human IL-2 (100 IU/ml). The frequency of CMV-specific CD8+ T cells was determined using the CMV pp65/HLA-A2 tetramer (MBL). CMV-specific CD8+ T cells, which were 65-98% tetramer positive, were used in the assays between days 12 and 16 after stimulation with the CMV peptide.

Co-culture study of human CMV-specific CP8 ⁺ T cells with human melanoma cell lines expressing human PVR: An in vitro co-culture assay with human CMV-specific CD8+ T cells was used to evaluate the effect of anti-human TIGIT antibodies on antigen-specific cytokine secretion. The target cell line used in the assay was an HLA-A2+ melanoma cell line, Mel624, stably transduced with a lentivirus containing human PVR DNA (System Biosciences). Stable pool of Mel624 cells overexpressing human PVR injected with CMV pp65 peptide at 0.0033 μg/ml or 0.001 μg/ml at 37° C. for 1 hour. The cells were then washed and plated at 50,000 cells/well. Hybridoma and phage-derived anti-human TIGIT antibodies, control mIgG1 antibodies or isotype hIgG4 antibodies or BM anti-human TIGIT antibodies were added at a concentration of 10 μg/ml. Human CMV-specific CD8+ T cells from three different donors, identified as donor 2, donor 4 and donor 210, were expanded according to the above protocol. 50,000 human CD8+ T cells were added to each well. Incubi co-cultures

- 80 040773 were washed at 37°C with 5% CO2 for 24 hours. After incubation, the plates were centrifuged at 1200 rpm for 1 min and the supernatant was collected. The amount of human interferon gamma (IFN-γ) in the co-culture supernatant was measured by flow cytometry using a cytometric granule assay (BD).

Co-culture study of human CMV-specific CD8 ⁺ T cells with melanoma cell lines expressing human PVR and human PVRL2 (CD 112): combined effect of anti-human TIGIT antibodies and CHA.7.518.1.H4 (S241P), anti-human PVRIG antibodies antigen-specific cytokine secretion was assessed using an in vitro co-culture assay with human CMV-specific CD8+ T cells, similar to the assay described above. The target cell line used in the assay was an HLA-A2+ melanoma cell line, Mel624, which stably expressed human PVR and human PVRL2, ligands for TIGIT and PVRIG, respectively, via lentiviral transduction (System Biosciences). Mel624 cells overexpressing human PVR and human PVRL2 were injected with CMV pp65 peptide at 0.0033 µg/mL or 0.001 µg/mL at 37°C for 1 h. Cells were then washed and plated at 50,000 cells/well. Hybridoma and phage-derived anti-human TIGIT antibody or BM anti-human TIGIT antibody was added to the culture in combination with CHA.7.518.1.H4 (S241P) or control hIgG4 isotype antibody at 10 μg/ml. Human CMV-specific CD8+ T cells from three different donors, identified as donor 4, donor 25 and donor 210, were expanded according to the above protocol. 50,000 human CD8+ T cells were added to each well. Co-cultures were incubated at 37° C. for 24 hours. After incubation, the plates were centrifuged at 1200 rpm for 1 min and the supernatant was collected. The amount of human interferon gamma (IFN-γ) in the co-culture supernatant was measured by flow cytometry using a cytometric granule assay (BD).

3. Results

Anti-human TIGIT antibodies enhance IL-2 signaling: The ability of hybridoma and phage-derived anti-human TIGIT antibodies to enhance IL-2 signaling was assessed using the human TIGIT luciferase reporter gene/human PVR Jurkat assay. In FIG. 60 and 62 depict the effect of 10 μg/L anti-human TIGIT phage or hybridoma antibodies on IL-2 signaling, respectively. Three phage-derived antibodies, CPA.9.027, CPA.9.049, and CPA.9.059, stably enhanced IL-2 signaling compared to the hIgG4 isotype control. In addition, all three phage antibodies induced IL-2 signaling more strongly than the BM anti-human TIGIT antibodies, BM26 and BM29. Five hybridoma-derived antibodies, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547 and CHA.9.560 also induced IL-2 signaling compared to the mIgG1 isotype control. Of note, five hybridoma antibodies induced similar IL-2 signaling compared to BM26 and BM29. The non-blocking human TIGIT antibody, CHA.9.543, slightly increased IL-2 signaling. To determine if the effect of anti-TIGIT antibodies is dose-dependent, analysis was performed using 10-point, 2-fold dilution series for each antibody, starting at 20 μg/ml (FIGS. 61 and 63). IL-2 signaling decreased in a dose-dependent manner for all eight anti-human TIGIT antibodies, as well as BM26 and BM29.

Anti-human TIGIT antibodies increase IFN-γ secretion from human CMV-specific CD8+ T cells: The ability of hybridoma and phage-derived anti-human TIGIT antibodies to modulate IFN-γ secretion was assessed by a CMV-specific T cell/Mel624 co-culture assay. In FIG. 64 shows the effect of human anti-TIGIT antibodies on IFN-γ secretion. Three phage antibodies, CPA.9.027, CPA.9.049, and CPA.9.059, increased IFN-γ secretion compared to media alone and the hIgG4 isotype control antibody. In addition, five hybridoma antibodies, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547 and CHA.9.560 also increased IFN-γ production compared to the mIgG1 isotype control antibody. Phage and hybridoma antibodies to TIGIT induced IFN-γ similarly to BM26 and BM29. As expected, the non-blocking human anti-TIGIT antibody, CHA.9.543, had no significant effect on IFN-γ secretion.

In FIG. 65 depicts the combined effect of anti-human TIGIT antibodies and CHA.7.518.1.H4 (S241P) on IFN-γ secretion. Three phage-derived antibodies, CPA.9.027, CPA.9.049 and CPA.9.059, and five hybridoma-derived antibodies, CHA.9.536, CHA.9.541, CHA.9.546, CHA.9.547 and CHA.9.560, including BM26, all enhanced IFN-γ secretion compared to their respective isotype control antibodies when administered alone or in combination with CHA.7.518.1.H4 (S241P). The non-blocking anti-human TIGIT antibody, CHA.9.543, resulted in a decrease in IFN-γ secretion compared to other anti-human TIGIT antibodies. Percentage increase in IFN-γ secretion; for each antibody compared to the corresponding isotype control antibodies are shown in FIG. 66. A synergistic effect was observed with combined treatment with anti-human TIGIT antibodies and CHA.7.518.1.H4 (S241P).

4. Summary and conclusions

Addition of Anti-Human TIGIT Antibodies to Human TIGIT/Human PVR Reporter Gene Assay

- 81 040773 century Jurkat resulted in a sustained, dose-dependent increase in IL-2 signaling. In addition, anti-human TIGIT antibodies increased IFN-γ secretion from human CMV-specific CD8+ T cells when co-cultured with human PVR Mel624 cells. IFN-γ secretion was further increased by anti-human TIGIT antibodies in combination with anti-human PVRIG antibody. Taken together, these data demonstrate that anti-human TIGIT antibodies can block TIGIT-mediated suppression of human T-cell activation, and T-cell activation is enhanced by co-blockade of both TIGIT and PVRIG.

P. Example 16 Anti-TIGIT Antibody Sorting Assay

1. Protocols

The experiments were carried out by Wasatch Microfluidics Inc. (Salt Lake City, UT) using Continuous Flow Microspotter (CFM) and IBIS MX96 SPR Imager (MX96 SPRi). The following anti-human TIGIT mAbs and human PVR-Fc variants were diluted to ~10 µg/mL in 10 mM sodium acetate, pH 5.0 and covalently immobilized using a standard amine crosslink on independent sites of a Biantensor Xantec 200M biosensor prism chip for 7 min. cycles using CFM:

1 СРА.9.009-Н4 17 СРА.9.081-Н4 l 32 SNA.9.555 2 SRA.9.011-H4 18 SNA.9.519 33 SNA.9.560 3 CPA.9.012-H4 19 SNA.9.521 34SNA.9.525 4 СРА.9.013-Н4 20 ССА.9.522 35 SNA.9.538 5 SRA.9.014-H4 21 SHA.9.527 36SNA.9.543 6 CPA.9.015-H4 22 SNA.9.528 37 SNA.9.553 7 CPA.9.018-H4 23 SNA.9.529 38 SHA.9.556 8 CPA.9.027-H4 24 SNA.9.535 39SNA.9.561 9 CPA.9.049-H4 25 CHA.9.536 40 VM8-H4 10 CPA.9.053-H4 26 SNA.9.541 41 VM9-H4 AND CPA9.057-H4 27 CHA.9.546 42 VM26-N4 12 CPA9.059-H4 28 CHA.9.547 43 VM29-N4 13 CPA9.064-H4 29 CHA.9.549 44MBSA43-M1 14 CPA9.069-H4 30 CHA.9.552 45 PVR-Fc М2А 15 CPA9.071-H4 31 CHA.9.554 46 Sino PVR-FC

CPA9.077-H4

BM8-H4 and BM9-H4 refer to (US2O15/O21697OA1, clones 10A7 and 1F4 reformatted as hIgG4), respectively. MBSA43-M1 is a mouse anti-human TIGIT IgG1 from eBioscience. The prismatic chip was then washed with 1X PBST for 3 minutes and then immediately loaded into an MX96 SPRi imaging device where excess N-hydroxysuccinimide ester was quenched with a 5 minute injection of 1M ethanolamine. Preliminary experiments included several injection cycles of 100 nM monomeric human TIGIT (Sino Biologicals, cat. no. 10917-H08H) on all immobilized mAbs for 4 min followed by regeneration to test antibody binding activity and best determine regeneration conditions by assessing reproducibility of binding TIGIT. These preliminary experiments demonstrated that the best reagent for regenerating most of the immobilized mAbs was a 30 second injection of 1/500 phosphoric acid. However, the immobilized PVR did not retain activity and therefore their blocking patterns were generated and sorted as solution-only analytes. In these preliminary and sorting experiments described below, all protein samples were prepared in running buffer that was degassed with HBST. An epitope-specific sandwich sorting protocol was performed where each mAb and PVR were administered via TIGIT, which was complexed with each immobilized mAb, to determine if the immobilized mAb blocked binding of the mAb to TIGIT in solution. For each cycle, 100 nM TIGIT was first infused with an excess of all immobilized mAb for 4 min, followed by an immediate infusion of competitive mAb or ligand for 4 min at 274 nM (binding site concentration). This was repeated with each mAb and PVR acting as a competitive test substance. Control cycles with running buffer instead of competitive protein were performed every 12 cycles for double comparison. All surfaces were regenerated after each cycle with a 30 second injection of 1/500 phosphoric acid. Sensorgram data were processed and presented using proprietary Wasatch software. An antibody pair was classified as having a common TIGIT-binding epitope if no binding was found with the introduction of a TIGIT-excessive competitor that had previously been complexed with the immobilized mAb. An antibody pair was classified as binding to different epitopes on TIGIT or sandwiching if injection of a competitive mAb showed binding to pre-bound TIGIT. Low or minimal competitor binding was classified as pro

- 82 040773 interstitial blocker. Hierarchical clustering of paired TIGIT blocking patterns for each mAb and ligand was performed using proprietary Wasatch software.

2. Results

Both PVR-Fc and 13 mAb either lost activity or could not be regenerated as ligands, so their blocking patterns were identified as analytes in solution only. MCAT CPA.9.014-H4 was not sorted because it did not bind to TIGIT. In FIG. 67 shows dendrogram clustering based on paired blocking patterns for each mAb and two PVR proteins. The vertical axis represents the statistical coefficient of similarity in blocking patterns. Wasatch Microfluidics used a cutoff factor of 5 for mAb clustering, which is indicated by the line in FIG. 67. For a strict definition of an epitope group, where only those mAbs (and PVRs) that exhibit the same blocking patterns, there are a total of 12 discrete groups. If blocking patterns that show only minimal differences are grouped together, four closely related mAb and PVR communities are identified. These communities are indicated by different shaded boxes at the bottom of FIG. 67. In FIG. 68 are grouped together with MAbs and PVRs that populate each discrete unique group (bin), with each group indicated by a black square. The gray squares surround all of the unique groups that make up each community of related blocking characters/patterns. MAb and PVR in Fig. 68 are listed with a numerical code that represents each protein in the dendrogram of FIG. 67.

Q. Example 17: Administration of Anti-PVRIG Antibodies to TIGIT Knockout Mice

Rationale and goals

To test whether TIGIT deletion in combination with mouse PVRIG blockade can enhance tumor growth inhibition and survival in a syngeneic mouse tumor model.

Protocols Animals

TIGIT knockout (KO) mice were obtained from Ozgene Pty LTD (Australia). Wild-type (WT) C57BL/6 mice (Ozgene) served as controls. TIGIT KO and C57BL/6 female mice were used, eight to eleven weeks old. All studies were approved by the Institutional Committee for Animal Care and Use at Tel Aviv University (Tel Aviv, Israel).

Tumor models in vivo

1x105 B16/Db-hmgp100 melanoma cells were inoculated subcutaneously (sc) into the right flank of C57BL/6 DT or TIGIT KO mice. Antibody treatment was started on the same day as tumor inoculation (day 0), with 7-10 mice per treatment group. The antibodies used were isotype controls of mouse IgG1 (clone MOPC-21 BioXcell) and mouse IgG1 against mouse PVRIG (clone 407, Compugen LTD). Antibodies were administered at 10 mg/kg by intraperitoneal injection, twice a week for 3 weeks. Tumor growth was measured with an electronic caliper every 2-3 days and described as 0.5 x W ² x L mm ³ (L is the length, W is the width of the tumor). Animals whose tumor size reached 2250 mm3 were anesthetized.

Statistical analysis

Two-way repeated measures ANOVA followed by two-way repeated measures ANOVA for selected group pairs was performed using JUMP (Statistical Discoveries™) software. Analysis of tumor growth measurements was performed by comparing tumor volumes measured on the last day when all study animals were still alive. Statistical differences in percent from tumor-free mice were determined using the Mantel-Cox log-rank test. P values <0.05 were considered significant. * p<0.05; ** p<0.01; *** p<0.001.

results

Inhibition of tumor growth in vivo after treatment with a blocking anti-mouse PVRIG antibody in TIGIT KO mice

We tested the in vivo efficacy of TIGIT deletion in combination with mouse PVRIG blockade in a syngeneic B16/Db-hmgp100 subcutaneous melanoma mouse tumor model. Treatment of tumor-bearing C57BL/6 DT mice with a blocking anti-mouse PVRIG antibody had little effect on tumor growth inhibition (TGI) compared to isotype treatment (17% TGI at day 11 and 8% TGI at endpoint, at 18- e days). The effect of the TIGIT deletion on tumor growth was not significant compared to the C57BL/6 DT control group (17% TGI at day 11 and 13% TGI at endpoint). However, when the TIGIT deletion was combined with anti-mouse PVRIG antibody (clone 407), significant TGI was evident (63% on day 11 and 49% TGI at the end point) (FIGS. 80A and 80B). Consistent with TGI, TIGIT KO mice treated with anti-mouse PVRIG antibody (clone 407) showed increased survival compared to the C57BL/6 DT control group, but statistical significance was not reached (FIG. 80C).

Summary and conclusions

The combination of TIGIT deletion and PVRIG blockade significantly reduced tumor growth in vivo, indicating that both TIGIT and PVRIG play an inhibitory role in this melanoma tumor model.

- 83 040773

These data suggest that co-targeting TIGIT and PVRIG may be another combination therapy that significantly enhances antitumor responses.

R. Example 18: PVRIG Antagonism Enhances T Cell Effector Function and Reduces Tumor Growth

Essay

Despite recent advances, most patients do not receive long-term therapeutic benefit from checkpoint inhibitors. PVRIG is a novel immunosuppressive receptor of the DNAM/TIGIT family and we demonstrate here the role of PVRIG in regulating antitumor responses. PVRIG binds to PVRL2 and shows significantly increased expression on tumor-infiltrating lymphocytes compared to lymphocytes from normal tissues. PVRIG antagonism enhanced human T cell activation, and the combination of PVRIG with PD-1 inhibitors or TIGIT further synergistically increased lymphocyte function. We then reviewed the role of PVRIG in preclinical tumor models. PVRIG ^- / ^- mice demonstrated significantly increased in vitro T cell activation and reduced MC38 tumor growth, which was mediated by increased CD8 effector function. The anti-PVRIG antagonist antibody significantly reduced tumor growth when combined with anti-PD-L1 or when tested in TIGIT ^−/− mice. Thus, we demonstrate that the PVRIG-PVRL2 pathway was induced in human cancers and that PVRIG-PVRL2 antagonistic interactions resulted in increased T cell function and decreased tumor growth.

Significance state

These data demonstrate that PVRIG is a promising target for cancer treatment and provides a rationale for testing the PVRIG inhibitor, CHA.7.518.1.H4 (S241P), as a novel cancer immunotherapy agent, either as monotherapy or in combination with TIGIT or PD1 blockade. .

Introduction

Increasing evidence suggests that endogenous immune responses are critical for modeling the initiation, progression, and suppression of cancer (1)(2). The immune status of patients, as well as the content of tumor infiltrating lymphocytes (TIL) in the tumor microenvironment (TME), are key predictors of not only cancer survival rates, but also how patients respond to therapy (3) (4). T cells are a key component of TILs that can elicit an antitumor response, and most antitumor immune responses ultimately rely on the functionality of effector lymphocytes. CD8 T cell enrichment in a patient's tumor TME, as well as other factors that bias the immune response to an effective CD8 T cell response, such as mutation load and Th1 polarized TME, are key predictors of a favorable antitumor immune response (5) (6) .

A key observation in many solid tumors is that effector T cells have an activated or depleted phenotype in TME (7). This indicates that although T cells within the TME have an initially visible cognate antigen, once activated and translocated into the tumor, they are subsequently unable to elicit an effective antitumor response. Pre-activated or depleted T cells are defined by increased surface expression of co-inhibiting receptors such as PD-1 and CTLA-4 (8). Therapeutically, targeting these co-inhibitory receptors with antibodies that inhibit interaction with their cognate ligands has demonstrated significant clinical efficacy in patients with multiple advanced cancers (9). It has been mechanically shown that targeting these inhibitory receptors leads to the expansion of already tumor-reactive T cells that previously existed in TME and to the formation of T cell pools with an expanded diversity of T cell receptors (10) (11) (12). Although checkpoint inhibitors have now revolutionized cancer treatment and demonstrated the power of the immune system in fighting cancer, many patients still relapse and/or do not respond to treatment. Therefore, a better understanding of the immune response in cancer and targeting additional immune-based pathways will lead to additional therapeutic options.

Among these new pathways, a group of receptors and ligands within the nectin family and nectin-like molecules are currently being explored as potential new cancer immunotherapies. Receptors in this family include DNAM-1 (CD226), CD96 (TACTILE), TIGIT and most recently PVRIG (CD112R) (13) (14) (15). Of these molecules, DNAM is the activating receptor in this subfamily, binding to 2 ligands, PVR (CD155) and PVRL2 (CD112), to deliver an activating signal to lymphocytes (16). Two receptors in this family have been shown to inhibit human lymphocyte function, TIGIT and more recently PVRIG (17) (18). TIGIT is reported to have a high affinity interaction with PVR, a significantly weaker affinity for PVRL2, and has been shown to inhibit both T cell and NK cell response by delivering an inhibitory signal to lymphocytes via its ITSM motif (19) ( 20). More recently, PVRIG has been shown to bind with high affinity to PVRL2 and deliver an inhibitory signal through its ITIM motif (15). In both

- 84 040773 cases, the affinity of TIGIT for PVR and PVRIG for PVRL2 is higher than that of DNAM for PVR or PVRL2, indicating that TIGIT and PVRIG may outcompete PVR and PVRL2 from DNAM, providing an indirect mechanism by which TIGIT and PVRIG may decrease T-cell function. Within this family, PVR is also a ligand for CD96. The function of CD96 has been reported to be inhibitory on mouse lymphocytes (21) but activation on human lymphocytes (22). Based on these data, we postulate that in human lymphocytes 2 receptors, TIGIT and PVRIG, bind with high affinity to PVR and PVRL2, respectively, to deliver inhibitory signals to attenuate T cell function.

Although one recent report showed that human PVRIG inhibits the T cell response, the role of PVRIG and PVRL2 in cancer immune surveillance is not well understood. In particular, the expression profile of this pathway in oncological diseases and the role of PVRIG in the regulation of antitumor responses of CD8 T cells are not given. Moreover, the functional characterization of the mouse PVRIG gene and the impact of disruption of the PVRIG-PVRL2 interaction in vivo in preclinical tumor models is not reported. In this context, we elucidated the role of PVRIG in cancer development by reporting on the expression profile of PVRIG and PVRL2 in cancer, as well as the impact of PVRIG antagonism in tumor cell co-culture studies and in preclinical tumor models. We demonstrate that PVRIG has a differentiated expression profile on T cell subpopulations compared to TIGIT or CD96 and that PVRIG and PVRL2 expression is induced in cancer compared to normal adjacent tissues. In numerous human in vitro assay systems, a high affinity anti-PVRIG monoclonal antagonist antibody (CHA.7.518.1.H4 (S241P)) improved T cell function, particularly in combination with an anti-TIGIT or anti-PD1 antibody. In addition, we report novel characterization of murine PVRIG using antagonistic antibodies or PVRIG-deficient mice and demonstrate that inhibition of PVRIG-PVRL2 interaction reduces tumor growth, which has the strongest effect when combined with PD-1 inhibition or TIGIT genetic deficiency. Taken together, these data demonstrate that PVRIG is a critical inhibitory receptor in the regulation of T-cell antitumor responses and supports the development of CHA.7.518.1.H4 (S241P) for clinical trials in cancer patients.

Materials and methods

Study of human peripheral blood and tumor expression

PBMCs from healthy human donors were obtained from Stanford University in accordance with the Declaration of Helsinki. The human tissues were provided by the Human Tissue Network, a resource supported by the National Cancer Institute. Human cancer tissues and corresponding normal adjacent tissues were dissociated into single cells according to the manufacturer's protocol (Miltenyi Biotec). Dissociated cells were analyzed by flow cytometry to express different targets on different cell subpopulations. For each target expression on a distinct cell subpopulation, the fold expression value was calculated by dividing the target CIF value by the isotype control CIF value. Other researchers may have obtained samples from these same tissue samples. Tumor type was determined based on analysis of the pathology report for each sample. For IHC studies, anti-PVRL2 antibody (HPA-012759, Sigma) and PD-L1 (Sp142, SpringBio) were used for microanalysis of stained tumors (Biochain Institute) using the conditions described in Supplementary Methods. Evaluation was performed by 2 independent reviewers on duplicated biopsies from the same tumor.

Creation and characterization of antibodies to PVRIG

Anti-human PVRIG and anti-mouse PVRIG antibodies were generated as described in Supplementary Methods. Briefly, binding specificity and affinity of antibodies were assessed by selective binding to PVRIG-modified cells without detectable binding to cells that do not express the gene. The antagonistic activity of these anti-PVRIG antibodies was determined using ELISA-based assays and flow cytometry, in which the interaction of PVRIG with PVRL2 was impaired. For characterization in cellular assays, antibodies were tested in several T-cell-target cell culture assay systems, consisting of target cells that express PVRL2 in culture with PBMCs or tumor-derived T cells. gp100-specific T cell lines were expanded from melanoma tumors as described above (23). CMVpp65-reactive T cells were expanded from healthy donor PBMCs (CTL immunospot) with CMVpp65 (495-503), IL-2 and IL-7 for 10 days. For combined studies, antibodies to PD-1, TIGIT and PVRIG were used at a concentration of 10 μg/ml. Cytokine concentrations in the conditioned media were determined using a cytometric granular analysis (CBA) method and FACS staining was performed as described in additional methods.

Characterization of Mouse PVRIG Expression and Function

Mouse PVRIG binding interactions with mPVRL2 and mPVR were assessed by SPR and ELISA using recombinant PVRIG, PVRL2 and PVR proteins and by FACS using ectopically modified cell lines overexpressing PVRIG and PVRL2 or cell lines transfected with PVR or PVRL2 siRNA. PVRIG- and TIGIT-deficient mice

- 85 040773 was obtained as described in additional methods. Expression analysis was performed to study the expression of PVRIG in the spleen, lymph node and tumor in various cell subpopulations. Cellular functional assays demonstrating T cell modulation activity for murine PVRIG were established using WT and PVRIG ^- / ^- T cells and target cells ectopically expressed by PVRL2 Fc or PVRL2, as described in Supplementary Materials and Methods. Tumor models CT26, MC38 and B16/Db-hmgp100 were performed as described in Supplementary Methods. All studies were approved by the Institutional Animal Care and Use Committee and Tel Aviv University (Tel Aviv, Israel) or Johns Hopkins University (Baltimore, USA).

results

PVRIG expression is highest on effector T cells in peripheral blood and tumors.

The Ig superfamily (IgSF) consists of hundreds of proteins, but only a few of them are inhibitors of T-cell receptors. IgSF proteins are rapidly evolving (24) and therefore sequence similarity among these proteins is usually low and not optimal for identifying new immune receptors. To identify new immune checkpoints, we developed bioinformatic algorithms based on common genomic and proteomic characteristics among known immune checkpoints, such as gene structure, protein domains, predicted cellular localization, and expression pattern. Using these algorithms, PVRIG has been identified as a novel immune receptor. A recent report also showed that human PVRIG (CD112R) binds to PVRL2 and inhibits T-cell function (15). However, the significance of this pathway in the regulation of tumor immune surveillance is not reported. In this context, we elucidated the expression and function of PVRIG and PVRL2 in human cancers and preclinical tumor models. In peripheral blood from healthy donors, PVRIG was exclusively expressed on lymphocytes, with the highest expression on CD8 T cells and NK cells (FIG. 83A). Further analysis of the T cell subsets demonstrated the highest expression of PVRIG on the CD8 or CD4 memory/effector T cell subsets compared to the regulatory T cell subset (FIG. 83B, FIG. 90A). The predominant pattern of memory T cell expression distinguishes PVRIG from other receptors in the family (TIGIT, CD96) which tend to have equal or higher expression on regulatory T cells compared to memory T cells/effector T cells. We additionally compared the expression kinetics of PVRIG and TIGIT activation after T cell activation in 2 assay systems (response to CMV stimulation in Fig. 83C, DC-MLR, Fig. 83D, Fig. 90B) and showed that PVRIG has a delay in induction kinetics and more stable expression at a later time point compared to TIGIT. The preferential expression of PVRIG on memory/effector cells over TIGIT points to a unique role for PVRIG in regulating T cell responses.

The delayed and sustained induction of PVRIG expression on T cells after activation suggests that it may be expressed in the tumor microenvironment. We then analyzed PVRIG expression on leukocytes from dissociated human tumors directly ex vivo using FACS. PVRIG expression was detected on CD8 T cells, CD4 T cells, and NK cells from several tumor types (Fig. 83E-G, Fig. 90C). PVRIG was co-expressed with PD1 and TIGIT on CD4 and CD8 T cells (FIG. 83F). On average, higher expression was found on CD4+ and CD8 ⁺ TIL from breast, endometrial, head and neck, lung, kidney, and ovarian tumors compared to bladder, colon, and prostate tumors. In tumor samples in which PVRIG expression was low/absent ex vivo, activation of anti-CD3 and anti-CD28 increased PVRIG expression, suggesting that PVRIG expression in TIL could be further induced upon reactivation (FIG. 90D). For colon, lung, kidney, endometrial, and ovarian tumors, we were able to obtain normal adjacent tissue from the same patient and compare PVRIG expression on lymphocytes isolated from the tumor versus normal tissue. TILs showed significant induction of PVRIG on CD4 and CD8 T cells compared to cells isolated from corresponding normal adjacent tissues (NAT) (FIG. 90E). As with PBMCs, we additionally compared PVRIG, TIGIT, and PD1 expression on regulatory T cells against CD8 T cells from lung, endometrial, and kidney tumors. At TIL, TIGIT expression was higher on regulatory T cells compared to CD8 T cells, while for PVRIG and PD1, similar or higher expression was observed on CD8 T cells than on regulatory T cells (Fig. 83H). We then examined the co-regulation of PVRIG, TIGIT and PD-1 on T cell populations by correlation analysis of either the magnitude of expression on TIL ex vivo or the magnitude of the fold change in expression between tumor and NAT. In both assays, CD4 and CD8 T cells showed a positive and significant correlation between PVRIG and PD1 or TIGIT (FIG. 90F). Taken together, these data demonstrate that PVRIG is expressed on T cells and NK cells in multiple human cancers, making PVRIG a novel target inhibitory receptor that may be critical for the regulation of T cell function in tumors.

PVRL2 expression is upregulated in tumor tissue compared to normal adjacent tissue.

- 86 040773

As PD-L1 expression has been shown to help predict responses to PD-1 inhibitors, we investigated whether PVRL2 expression was associated with expression of its cognate receptor PVRIG in human cancer tissues. Using the anti-PVRL2 antibody that we tested for staining FFPE specimens (Fig. 91A), we stained tumor tissue matrices (TMAs) consisting of lung, colon, skin, breast, ovarian/endometrial, and kidney tissues and evaluated each biopsy. based on the prevalence and intensity of PVRL2 expression. PVRL2 expression was absent or minimally expressed in most normal tissue samples from these organs. In tumor tissues, PVRL2 expression on tumor cells was found in ~50-70% of lung, colon, breast, and ovarian/endometrial cancers (FIGS. 84A, 84F). Expression in kidney cancer samples ranged from 20 to 40%, while expression in melanoma was the lowest (~10%) (Fig. 84A, 84F). PVRL2 expression was found on tumor cells and immune cells at the invasive front (FIG. 84B). To identify specific subpopulations of immune cells expressing PVRL2, we performed flow cytometry on freshly dissociated tumors. PVRL2 expression was detected on CD45+ immune cells, in particular myeloid cells (eg, CD14+ tumor-associated macrophages (TAMs) and myeloid DCs) and on CD45-non-immune cells from several tumor types (Fig. 84C,D). No PVRL2 expression was found on lymphocytes (data not shown). Comparison of PVRL2 expression on CD45 cells and TAM isolated from colon, lung, kidney, endometrial, and ovarian tumors demonstrated a significant induction of PVRL2 on cells isolated from the tumor compared to cells isolated from the corresponding NAT from the same donor (Fig. .92). For samples where we obtained PVRIG and PVRL2 expression, we examined PVRIG expression on lymphocytes versus PVRL2 on myeloid and ^CD45 cells from several tumor types. Of the cancer types studied, endometrial, lung, and kidney tumors had the highest prevalence of PVRIG ^hl lymphocytes and PVRL2 ^hi TAM or CD45 non-immune cells (Fig. 842E, 93). By integrating data from studies of TMA and dissociated tumors, we demonstrate that tumors of the breast, endometrium, lung, head and neck, kidney, and ovary may represent a tumor type susceptible to PVRIG antagonism.

Compared to PD-L1, PVRL2 expression is differentially regulated and is present in PD-L1 tumors

Since PVRIG and PD-1 can be co-expressed on tumor infiltrating lymphocytes (TILs), we also examined the co-expression of PVRL2 and PD-L1 on the same tumor by staining consecutive sections with the same TMA. Expression of PVRL2 on tumor cells was clearly detected in PD-L1' tumor samples (as defined by the absence of PD-L1 membrane staining on tumor cells or immune cells) at the same frequency and mean as compared to PD-L1+ samples. (FIGS. 85A, 84F). On immune cells, 3 out of 5 tumors in which PVRL2 expression was found on immune cells also express PD-L1 (data not shown), but the small number of samples makes it difficult to conclude that PD-L1 and PVRL2 are co-expressed in the immune cell. Expression of PVRL2 on tumor cells in PD-L1-negative tumors demonstrated that PVRL2 expression was more abundant than PD-L1 in some tumor types and that targeting this pathway may be particularly effective in PD ^- L1 tumors. While PD-L1 is induced primarily by IFN-γ as an adaptive resistance mechanism (28), PVRL2 is modulated by genomic stress, DNA damage, and tumor suppressor genes (29, 30). To further understand the precise regulation of PDL1 and PVR/PVRL2, we next assessed the regulation of PVR, PVRL2 and PD-L1 expression in tumor cell lines and in monocytic DCs by exposure to various inflammatory stimuli (Fig. 85D). Treatment of DCs with pro-inflammatory signals generally results in increased expression of PVR, PVRL2, and PD-L1, demonstrating that PVR, PVRL2, and PD-L1 expression increase after DC maturation. In contrast, treatment with epithelial cells with IFN-γ increased PD-L1 expression but had no effect on high baseline PVRL2 expression (Fig. 85E), maintaining differential regulation of PVRL2 expression compared to PD-L1 by IFN-γ. Thus, these data demonstrate that PD-L1 and PVRL2 can be co-regulated on antigen presenting cells (APCs) such as DCs, but can be differentially regulated on epithelial cells. The presence of PVRL2- in PD-L1-negative tumors indicates that targeting this pathway may be potentially beneficial in patients who do not respond to or progress on PD-1 inhibitors.

CHA.7.518.1.H4 (S241P) is a high affinity humanized anti-PVRIG monoclonal antibody that interferes with the interaction of PVRIG with PVRL2.

To investigate the functional implications of human PVRIGPVRL2 antagonistic interactions, we created a high affinity anti-PVRIG antagonist antibody, CHA.7.518.1.H4 (S241P), which blocks the interaction between PVRIG and PVRL2. This antibody selectively binds HEK293 cells ectopically expressing human PVRIG or Javanese PVRIG.

- 87 040773 kaka, and also bound Jurkat cells that endogenously express PVRIG with subnanomolar affinity (Fig. 86A). In biochemical assays, CHA.7.518.1.H4 (S241P) blocked the interaction of PVRIG Fc with PVRL2+ HEK293 cells (FIG. 86B) and also blocked the binding of PVRL2 Fc to PVRIG+ HEK293 cells (FIG. 86C). Using this antibody, we observed the functional effect of antagonistic anti-PVRIG in several T cell assays. Artificial antigen presenting cells (aAPCs) ectopically expressing anti-human CD3 and PVRL2 antibodies were generated and co-cultured with primary human CD4 T cells or in the presence of anti-PVRIG (CHA.7.518.1.H4(S241P)) or isotypic control. PVRIG expression was induced on proliferating CD4 T cells when co-cultured with CHO anti-CD3 aAPC (FIG. 94A). Antagonism of PVRIG with CHA.7.518.1.H4 (S241P) enhanced CD4T cell proliferation from multiple donors (FIG. 86D). We also tested the effect of anti-PVRIG on 2 gp100 reactive human CD8 T cell lines derived from melanoma tumors. These T cell lines were individually co-cultured with HLA-A2 and PVRL2 expressing aAPCs (FIG. 94B) in the presence of IgG isotype control or anti-PVRIG antibodies. As noted in both lines, anti-PVRIG increased the production of IFN-γ and TNF-α by ~20-50%. In dose-response evaluation, CHA.7.518.1.H4 (S241P) displayed unambiguous nanomolar EC50 values in several assays (Fig. 94C,D). These data taken together demonstrate that antagonistic interactions of PVRIG-PVRL2 with CHA.7.518.1.H4 (S241P) resulted in increased T cell activation.

CHA.7.518.1.H4 (S241P) in combination with TIGIT or PD-1 inhibitors resulted in a synergistic increase in T cell function.

The combination of PVRIG and TIGIT blockade synergistically increased CD4 T cell function in a T cell-dendritic cell coculture assay (15), suggesting a role for this pathway in regulating T cell-APC interactions. The effects of PVRIG and TIGIT blockade on CD8 T cells under co-culture conditions of tumor cells have not been reported. Because our tumor expression profile demonstrated PVRL2 expression on CD45 immune cells, we further examined the effect of targeting this pathway in T-tumor cell co-culture using 2 T-cell assay systems. We first co-cultured 2 gp100 neoantigen-specific CD8 T cell lines with a melanoma cell line, MEL624, in the presence of anti-PVRIG, anti-TIGIT, or isotype antibodies, either alone or in combination. MEL624 cells express both PVR and PVLR2, and TIL-209 and TIL-463 express PVRIG, TIGIT and PD1 (FIG. 86F). At TIL-209, we noticed that anti-PVRIG or anti-TIGIT alone did not increase IFN-γ and that the combination of anti-PVRIG and anti-TIGIT synergistically increased IFN-y production (FIG. 86G). At TIL-463, we observed that anti-PVRIG or anti-TIGIT modestly increased IFN-γ production and that the combination of anti-PVRIG and anti-TIGIT additively increased IFN-γ (FIG. 86G). In an additional assay system, we used CMVpp65-reactive CD8 T cells as a model system to study human T cell responses. HLA-A2+ CMVpp65 CD8 T cells were expanded in the presence of CMVpp65 (495-503) and PVRIG, TIGIT and PD-1 expression was observed on day 10 (FIG. 86F). PVRIG was expressed on CMVp65-specific CD8 T cells with the same magnitude as in human cancer samples (FIG. 83). As target cells, we identified PD-L1 ^hl (Panc05.04) and PD-L1 ^|0 (Colo205) HLA-A2 ⁺ cancer cell lines that expressed similar amounts of PVR and PVRL2 (Fig. 86F). We then co-cultured CMVpp65-reactive T cells with HLA-A2 ⁺ tumor cell lines injected with pp65(495-503) peptide in the presence of blocking antibodies to PVRIG, TIGIT and/or PD-1. We observed that anti-PVRIG Ab increased IFN-γ by ~50% in co-culture with Pap05.04 cells and minimally in co-culture with Colo205 (FIG. 861). The combination of anti-TIGIT with anti-PVRIG Ab synergistically increased IFN-γ production in both target cell lines resulting in a greater increase in IFN-γ compared to anti-PD-1 alone (FIG. 86H). The combination of anti-PVRIG and anti-PD-1 also resulted in a synergistic increase in IFN-γ production compared to a single antibody (FIG. 861). Taken together, these data indicate a strong synergy of combining PVRIG and TIGIT or PVRIG and PD1 blockade in increasing human CD8 T cell activation when interacting with tumor cells.

PVRIG deficiency resulted in increased T cell proliferation and decreased tumor growth.

Although the sequences for mouse PVRIG and its interaction with mouse PVRL2 have been described, the expression profile and immunomodulatory activity of mouse PVRIG is not well understood. We first analyzed RNA expression and mPVRIG transcript in NK, NKT and T cells (Fig. 87A). Activated mouse CD8 T cells had increased PVRIG transcripts with delayed induction kinetics compared to TIGIT (FIG. 87B). We confirmed that recombinant mPVRIG bound to the mPVRL2 protein using surface plasmon resonance (SPR) and ELISA, which was performed in multiple assay orientations (Fig. 95A-D). We also observed an interaction between mPVRIG and mPVR, although the affinity was about 10 times less than the interaction with mPVRL2 (Fig. 95E). To determine if PVR or PVRL2 is dominant

We tested the binding of mouse PVRIG Fc to B16F10 cells that express PVR and PVRL2 (data not shown). PVRIG Fc showed dose-dependent binding to B16F10 cells, which was completely abolished when PVRL2 siRNA was knocked down in B16F10 cells (FIG. 95F). In comparison, PVRIG Fc fusion protein binding was not significantly but consistently decreased after PVR knockdown (FIG. 95F), indicating little interaction between mPVRIG and mPVR. Taken together, these results demonstrate that in mice, PVRL2 is the main ligand for PVRIG, as is the case in humans.

To determine the role of PVRIG in immune responses, we generated PVRIG-deficient ( ^-/- ) mice (FIG. 96). PVRIG ⁻ / ⁻ mice were born at expected Mendelian ratios, did not show a pronounced phenotype until 10 months of age, and at 8 weeks of age had similar leukocyte saturation (peripheral and lymphoid tissue) compared to wild-type mice (Fig. 97). CD8 wild-type (WT) and NK cells express PVRIG and no PVRIG expression was detected on PVRIG ^- / ^- cells (Fig. 87C). To investigate the role of PVRIG in regulating mouse T cell responses, we examined the proliferation of DT and PVRIG ^-/- T cells in 2 assay systems. DT or PVRIG ^−/ T cells were activated with immobilized anti-CD3 in the presence of soluble PVRL2 Fc protein or control Fc protein. Soluble PVRL2 Fc significantly inhibited the proliferation of CD4 ⁺ T- cells of TD, but not the proliferation of PVRIG ^- / ^- CD4 + T cells (Fig. 87D), indicating that PVRIG ^- / ^- cells do not have an inhibitory signal. To assess the role of mouse PVRIG in interaction with CD8+ T cells with tumor cells, PVRIG ⁻ / ⁻ mice were crossed with pmel TCR transgenic mice that expressed a transgenic TCR specific for gp10025-33 (28). Activated PVRIG ^- / ^- or DT Pmel CD8+ T cells were co-cultured with B16-Db/gp100 melanoma tumor cells that endogenously expressed PVRL2 (data not shown) and evaluated for activation and effector function. PVRIG ^- / ^- pmel CD8 ⁺ T cells showed enhanced degranulation and production of effector cytokines (IFN-γ and TNF-α) compared to DT cells (Fig. 87E). These data indicate that mouse PVRIG inhibits the activation and effector function of tumor-specific T cells when co-cultured with PVRL2+ target tumor cells.

We then examined the effect of PVRIG deficiency on tumor growth in a syngeneic MC38 model. PVRIG^-/ mice showed significantly reduced tumor growth compared to DT mice (p<0.05; FIG. 88AB). In addition, PD-L1 blockade started on day 14 additionally enhanced antitumor responses and reduced tumor growth in PVRIG.^-/- mice compared to DT mice treated with anti-PD-L1 (p = 0.052) (Fig. 88C-D). To evaluate the functional effects of PD-L1 blockade on the microenvironment of PVRIG tumors^-/^- and WT, we collected tumors and tumor-draining lymph nodes from each of the four experimental cohorts on day 18, when the groups received 2 doses of either isotype or anti-PD-L1, but found no difference in tumor volume, and performed flow cytometry on the composition of the immune subpopulation and intracellular cytokines. Migration of immune cells (CD45+) to PVRIG^-/^- tumors moderately increased (88% compared with TD tumors), both for CD8+ T cells (92% compared with TD tumors) and for IFN-γ-producing CD8+ T cells (110% increase compared with tumors DT, Fig. 88E). In combination with PD-L1 blockade, CD45+ cell infiltration was significantly increased in PVRIG.^-/^- tumors (160% relative to tumors from TD mice treated with PD-L1, W = 0.032, Fig. 88F). At PVRIG^-/^- tumors treated with anti-PD-L1 also had higher numbers of CD8+ T cells per tumor mass (up to 252% increase) and CD8⁺ T-lymphocytes producing IFN-γ (297%) compared with anti-PD-L1 treated TD tumors (Fig. 88F). We also noticed that PVRIG^-/- mice had unchanged numbers of effector tumor-infiltrating CD4+ T cells and Foxp3⁺ regulatory T cells regardless of PD-L1 blockade (data not shown). Relief from immune dysfunction in PVRIG^-/- tumors, especially after PD-L1 blockade, was reflected in tumor-draining lymph nodes, which had an increased occurrence of IFN-γ⁺ TNF-α⁺ effector CD8⁺ T cells versus anti-PD-L1 treated TD mice (Fig. 88G-H). Taken together, these data demonstrate that the removal of PVRIG results in a reduction in tumor growth associated with an increased anti-tumor immune response, particularly in combination with an anti-PD-L1 antibody.

An anti-mPVRIG antibody inhibits tumor growth in combination with an anti-PD-1 antibody or TIGIT deficiency.

After demonstrating that PVRIG genetic deficiency leads to reduced tumor growth, we further sought to demonstrate that antibody-mediated inhibition of the PVRIG-PVRL2 interaction can improve antitumor immunity, particularly in combination with PD1 or TIGIT inhibitors, as demonstrated by our data in vitro for humans. To evaluate this, we created a high affinity anti-mPVRIG antagonist antibody. Anti-mPVRIG mAb affinity scores determined by FACS showed a subnanomolar Kd (0.33 nM on HEK293 mPVRIG, 0.39 nM on D10.G4.1 cells) similar to CHA.7.518.1.H4 (S241P) (Fig. .95GH). The specificity of this antibody was further confirmed by the fact that most of the binding to D10.G4.1 cells was impaired by mPVRIG knockdown (FIG. 95I). Anti-mPVRIG tested

- 89 040773 to disrupt mPVRG-mPVRL2 interaction by inhibiting mPVRG Fc binding to B16F10 and mPVRL2 Fc binding to mPVRIG-overexpressing HEK293 cells (FIG. 89A). Complete blocking of the PVRIG-PVRL2 interaction with the anti-mPVRIG antibody was observed in both assay formats (FIG. 89A, FIG. 95J), demonstrating an anti-mPVRIG antagonistic antibody. Next, we tested the in vivo efficacy of mPVRIG blockade in a syngeneic CT26 subcutaneous colon tumor model. PVRIG expression was increased on NK and T cells in the tumor microenvironment compared to the corresponding subpopulations from the spleen or draining lymph nodes (FIG. 89B). Treatment of tumor-bearing mice with anti-mPVRIG blocking mAb as monotherapy failed to reduce tumor growth (data not shown). However, the combination of anti-PVRIG and antiPD-L1 mAbs effectively delayed CT26 tumor growth (FIG. 89C) and significantly increased the survival rate of treated mice by 40% of those with complete response (FIG. 89D). Consistent with our analysis of human T cells, these data demonstrate that the combination of PD-1 and PVRIG inhibitors can reduce tumor growth.

We also tested the effect of removing both PVRIG and TIGIT signals in regulating antitumor responses. For these studies, we tested the efficacy of anti-mPVRIG antibodies in DT or TIGIT ^- / ^- mice inoculated with B16F10/Db-hmgp100 melanoma cells. Treatment of tumor-bearing TD mice with an anti-mPVRIG blocking mAb had little effect compared to isotype treatment (17% TGI at day 11 and 8% TGI at the end point, day 18). The effect of the TIGIT deletion on tumor growth was also insignificant compared to the DT control group (17% TGI at day 11 and 13% TGI at the endpoint). However, when TIGIT deletion was combined with anti-PVRIG mAb treatment, significant tumor growth inhibition was observed (63% at day 11 and 49% TGI at the endpoint (Fig. 89E, F). Consistent with TIGIT tumor growth inhibition ^- / ^- mice treated with anti-PVRIG mAb 407 showed increased survival compared to DT controls, however statistical significance was not reached in this aggressive fast growing tumor model (data not shown) Collectively, these data demonstrate synergistic activity of PVRIG inhibitors with inhibitors PD1 or TIGIT and are consistent with our human functional data, which provides the basis for clinical trials of CHA.7.518.1.H4 (S241P) with PD1 or TIGIT inhibitors.

Discussion

Although antibodies targeting immune T cell checkpoints, such as CTLA4 and PD-1, have increased cancer patient survival, most cancer patients still do not show clinical benefit. One possible reason for this is the presence of additional T cell regulators that inhibit T cell antitumor immunity. In this paper, we elucidated the role of PVRIG in the regulation of T cell effector function and demonstrated that PVRIG antagonism increases T cell antitumor responses and reduces tumor growth.

PVRIG is a new member of the family of nectins and nectin-like molecules, placing it among the few known immunoregulatory receptors in the family. Understanding the interaction of receptors within this family is critical to understanding the relevance and mechanism of action of PVRIG. Of these DNAM receptors, TIGIT and CD96 are most closely related to PVRIG in terms of sharing the same ligands, PVR and PVRL2. DNAM binds to both PVR and PVRL2 and provides a costimulatory signal to lymphocytes. TIGIT is reported to bind to PVR and loosely binds to PVRL2. We were unable to detect an interaction between TIGIT and PVRL2 using ELISA or SPR (data not shown), suggesting that PVR is the dominant ligand for TIGIT. Using similar methods, the latest report found a high affinity interaction between PVRL2 and PVRIG, indicating that PVRIG is the dominant inhibitory PVRL2 receptor. These data suggest that TIGIT and PVRIG contain dual signaling nodes on this axis and that blocking both is necessary to maximize T cell activation in this family. In addition to interacting with various ligands, we have observed that PVRIG has the highest expression on effector or memory T cells, similar to PD-1, while TIGIT has the highest expression on regulatory T cells. In addition, we observed that PVRIG exhibited late induction after T cell activation compared to TIGIT. These data suggest that PVRIG has a unique role in this family, interacting with high affinity for PVRL2 and having differential expression on memory cells and a late induction profile for TIGIT.

A novel role for PVRIG in the regulation of anti-tumor T-cell responses is reported using PVRIG-deficient mice and antagonistic anti-PVRIG antibodies. In this document, it was demonstrated that mouse PVRIG is expressed on T cells and NK cells induced by lymphocyte activation and is highest in TME compared to peripheral ones. In addition, we demonstrate that PVRIG deficiency resulted in an increase in T cell function in vitro and a decrease in tumor growth in vivo. An anti-PVRIG antagonist antibody reduced tumor growth in combination with anti-PD-L1 or TIGIT genetic deficiency, demonstrating a necessary role for PVRIG in regulating T cell responses. These new data provide an in vivo proof of concept using

- 90 040773 using preclinical tumor models that targeting PVRIG in combination with PD1 antagonism or TIGIT is a potential new therapy for cancer treatment.

This document provides information on a high affinity anti-human PVRIG antibody that interferes with the interaction between PVRIG and PVRL2, which we are testing in clinical trials. To identify potential cancer features that may inform patient choice in clinical trials, we analyzed the expression profile of this axis in human cancers using FACS and IHC. For PVRIG, we observed that mean PVRIG expression on CD4 and CD8 T cells by FACS is highest in endometrial, lung, kidney, and ovarian cancers, although this difference did not reach statistical difference with other cancer types as determined by variance analysis. analysis with Tukey's multiple comparison test with the current number of samples. Because PVRIG is induced upon T cell activation, and given that most tumor-infiltrating T cells have already encountered the antigen, it is perhaps not surprising that median PVRIG expression was similar between tumor samples and cancer types. We observed that PVRIG expression correlated with PD-1 and TIGIT expression, suggesting that the interaction of these 3 inhibitory receptors would be important in regulating the antitumor response. In this report, we observed a synergistic increase in T cell function when anti-PVRIG antibodies were combined with anti-TIGIT antibodies in co-cultured tumor cells and CD8 T cells, better than PD-1 in combination with PVRIG or TIGIT inhibitors. These data, together with previous research demonstrating the role of PVRIG and TIGIT in the regulation of DC-T cell interactions, demonstrate that this pathway may be involved in the regulation of APC T-cell and T-tumor cell interactions and provides multiple mechanisms by which targeting PVRIG may increase the antitumor immune response.

Because PD-L1 expression correlated with clinical response to PD-1 inhibitors, we also analyzed PVRL2 expression in tumors using FACS and IHC to assess whether certain types of cancer are overexpressed. By evaluating dissociated tumor cells, we found that mean PVRL2 expression on macrophages from endometrial, head and neck, kidney, lung, and ovarian samples was higher compared to other tumor types. Mean PVRL2 expression on CD45 non-immune cells was higher in breast, colon, endometrial, lung, ovarian, and prostate cancers compared to other cancers. Based on the expression of PVRIG and PVRL2, we determined that cancers of the endometrium, head and neck, lung, kidney, and ovary are at greater risk of developing tumors with high expression of PVRIG and PVRL2 and that these are potential cancers that may respond to inhibitors of this pathway.

This document noted that PVRL2 expression can be modulated on antigen producing cells in vitro by inflammatory mediators, while PVRL2 expression on cancer cells was not altered. These data demonstrate that PVRL2 expression on antigen presenting cells may be regulated by inflammation and may be indicative of an inflamed tumor. Indeed, we have observed that all PD-L1+ tumors also express PVRL2 both on tumor cells and in the immune system compartment. Expression of PVRL2 on myeloid cells may help predict responses to PVRIG inhibitors in combination with PD-1 or TIGIT to further enhance the antitumor effect. Interestingly, a subset of PD-L1-negative tumors also expressed PVRL2, primarily on tumor cells rather than immune cells. It is reported that the expression of PVR and PVRL2 on epithelial cells is induced during tumor formation, as well as in response to stress and DNA damage. These data are consistent with in vitro findings that the regulation of PVRL2 expression on tumor cells is independent of IFN-gamma. Because PD-L1 is induced in adaptive resistance in response to IFN-gamma and is associated with an inflammatory response, PVRL2 expression in the absence of PD-L1 suggests that PVRL2 expression is more common than PD-L1 and that PVRL2 is expressed in non-inflammatory tumors. Based on the above, it is possible that the presence of PVR and PVRL2 contributes to the suppression of immune responses independent of PD-L1 and that inhibitors of PVRIG and TIGIT may be of particular importance in patients who are PD-L1 negative or do not respond/progress on PD-1 inhibitors. .

Thus, this report provides several new insights into the biology of PVRIG, including characterization of the expression of this axis in human cancers, demonstrating the important role of PVRIG/TIGIT in regulating CD8-tumor cell interactions, and demonstrating that PVRIG antagonism in combination with PD is 1 or TIGIT deficiency lead to a synergistic reduction in tumor growth. These data expand our current understanding of the biology of PVRIG and provide a rationale for clinical testing of CHA.7.518.1.H4 (S241P), a high affinity anti-PVRIG antibody, in patients with cancer.

- 91 040773

Bibliography

1. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100(l):57-70.

2. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144(5):646-74 doi 10.1016/j.cell.2011.02.013.

3. Galon J, Mlecnik B, Bindea G, Angell HK, Berger A, Lagorce C, et al. Towards the introduction of the 'Immunoscore' in the classification of malignant tumours. J Pathol 2014;232(2): 199-209 doi 10.1002/path.4287.

4. Zitvogel L, Galluzzi L, Smyth MJ, Kroemer G. Mechanism of action of conventional and targeted anticancer therapies: reinstating immunosurveillance. Immunity 2013;39(l):74-88 doi 10.1016/j.immuni.2013.06.014.

5. Danilova L, Wang H, Sunshine J, Kaunitz GJ, Cottrell TR, Xu H, et al. Association of PD-l/PD-L axis expression with cytolytic activity, mutational load, and prognosis in melanoma and other solid tumors. Proc Natl Acad Sci U S A 2016;113(48):E7769-E77 doi 10.1073/pnas. 1607836113.

6. Topalian SL, Taube JM, Anders RA, Pardoll DM. Mechanism-driven biomarkers to guide immune checkpoint blockade in cancer therapy. Nat Rev Cancer 2016;16(5):275-87 doi 10.1038/nrc.2016.36.

7. Zarour H.M. Reversing T-cell Dysfunction and Exhaustion in Cancer. Clin Cancer Res 2016;22(8): 1856-64 doi 10.1158/1078-0432.CCR-15-1849.

8. Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 2012;12(4):252-64 doi 10.1038/nrc3239.

9. Sharma P, Allison JP. Immune checkpoint targeting in cancer therapy: towards combination strategies with curative potential. Cell 2015;161(2):205-14 doi 10.1016/j.cell.2015.03.030.

10. Cha E, Klinger M, Hou Y, Cummings C, Ribas A, Faham M, et al. Improved survival with T cell clonotype stability after anti-CTLA-4 treatment in cancer patients. Sci Transl Med 2014;6(238):238ra70 doi 10.1126/scitranslmed.3008211.

11. Robert L, Tsoi J, Wang X, Emerson R, Hornet B, Chodon T, et al. CTLA4 blockade broadens the peripheral T-cell receptor repertoire. Clin Cancer Res 2014;20(9):2424-32 doi 10.1158/1078-0432.CCR-13-2648.

12. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, et al. PD1 blockade induces responses by inhibiting adaptive immune resistance. Nature 2014;515(7528):568-71 doi 10.1038/naturel3954.

- 92 040773

13 Chan CJ, Andrews DM, Smyth MJ. Receptors that interact with nectin and nectinlike proteins in the immunosurveillance and immunotherapy of cancer. Curr Opin Immunol 2012;24(2):246-51 doi 10.1016/j.coi.2012.01.009.

14. Martinet L, Smyth MJ. Balancing natural killer cell activation through paired receptors. Nat Rev Immunol 2015;15(4):243-54 doi 10.1038/nri3799.

15. Zhu Y, Paniccia A, Schulick AC, Chen W, Koenig MR, Byers JT, et al. Identification of CD112R as a novel checkpoint for human T cells. J Exp Med 2016;213(2):16776 doi 10.1084/jem.20150785.

16. Bottino C, Castriconi R, Pende D, Rivera P, Nanni M, Carnemolla B, et al. Identification of PVR (CD 155) and Nectin-2 (CD 112) as cell surface ligands for the human DNAM-1 (CD226) activating molecule. J Exp Med 2003;198(4):557-67 doi 10.1084/jem.20030788.

17. Yu X, Harden K, Gonzalez LC, Francesco M, Chiang E, Irving B, et al. The surface protein TIGIT suppresses T cell activation by promoting the generation of mature immunoregulatory dendritic cells. Nat Immunol 2009;10(l):48-57 doi 10.1038/ni.l674.

18. Stanietsky N, Simic H, Arapovic J, Toporik A, Levy O, Novik A, et al. The interaction of TIGIT with PVR and PVRL2 inhibits human NK cell cytotoxicity. Proc Natl Acad SciU S A2009; 106(42): 17858-63 doi 10.1073/pnas.0903474106.

19. Johnston RJ, Comps-Agrar L, Hackney J, Yu X, Huseni M, Yang Y, et al. The immunoreceptor TIGIT regulates antitumor and antiviral CD8(+) T cell effector function. Cancer Cell 2014;26(6):923-37 doi 10.1016/j.ccell.2014.10.018.

20. Zhang B, Zhao W, Li H, Chen Y, Tian H, Li L, et al. Immunoreceptor TIGIT inhibits the cytotoxicity of human cytokine-induced killer cells by interacting with CD 155. Cancer Immunol Immunother 2016;65(3):305-14 doi 10.1007/s00262-016-1799-4.

21. Chan CJ, Martinet L, Gilfillan S, Souza-Fonseca-Guimaraes F, Chow MT, Town L, et al. The receptors CD96 and CD226 oppose each other in the regulation of natural killer cell functions. Nat Immunol 2014; 15(5):431-8 doi 10.1038/ni.2850.

22. Fuchs A, Celia M, Giurisato E, Shaw AS, Colonna M. Cutting edge: CD96 (tactile) promotes NK cell-target cell adhesion by interacting with the poliovirus receptor (CD155). J Immunol 2004;172(7):3994-8.

23. Machlenkin A, Uzana R, Frankenburg S, Eisenberg G, Eisenbach L, Pitcovski J, et al. Capture of tumor cell membranes by trogocytosis facilitates detection and isolation of tumor-specific functional CTLs. Cancer Res 2008;68(6):2006-13 doi 10.1158/0008-5472, CAN07-3119.

24 Ohtani H, Nakajima T, Akari H, Ishida T, Kimura A. Molecular evolution of immunoglobulin superfamily genes in primates. Immunogenetics 2011;63(7):417-28 doi 10.1007/s00251-011-0519-7.

25. Taube JM, Anders RA, Young GD, Xu H, Sharma R, McMiller TL, et al. Colocalization of inflammatory response with B7-hl expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape. Sci Transl Med 2012;4(127):127ra37 doi 10.1126/scitranslmed.3003689.

26. Cerboni C, Fionda C, Soriani A, Zingoni A, Doria M, Cippitelli M, et al. The DNA Damage Response: A Common Pathway in the Regulation of NKG2D and DNAM-1 Ligand Expression in Normal, Infected, and Cancer Cells. Front Immunol 2014;4:508 doi 10.3389/fimmu.2013.00508.

27. de Andrade LF, Smyth MJ, Martinet L. DNAM-1 control of natural killer cells functions through nectin and nectin-like proteins. Immunol Cell Biol 2014;92(3):237-44 doi 10.1038/icb.2013.95.

28. Overwijk WW, Tsung A, Irvine KR, Parkhurst MR, Goletz TJ, Tsung K et algplOO/pmel 17 is a murine tumor rejection antigen: induction of "self"-reactive, tumoricidal T cells using high-affinity, altered peptide ligand . J Exp Med 1998; 188(2):277-86.

S. Example 19: Cell Tumor Kill Assay

The effect of anti-human TIGIT antibody and CHA.7.518.1.H4 (S241P), either alone or in combination, on tumor cell killing was assessed by in vitro co-culture assay with human CMV-specific CD8 ⁺ T cells. HLA-A2+ target cell lines, used

- 93 040773 used in the analysis were the melanoma cell line Me1624, which stably expresses human PVR and PVRL2, and the pancreatic adenocarcinoma cell line Raps05.04, which expresses endogenous levels of human PVR and PVRL2. Both tumor cell lines were stably transduced with the luciferase reporter gene by lentiviral transduction (System Biosciences). CMV pp65 peptide was injected into Mel624 and Raps05.04 cells at 0.0033 µg/ml or 0.01 µg/ml at 37°C for 1 h, respectively. The cells were then washed and plated at 20,000 cells/well. Reference antibodies against human TIGIT and CHA.7.518.1.H4 (S241P) were added to the culture in combination with or with the isotype control hIgG4 antibody at 10 μg/ml. Human CMV-specific CD8+ T cells from three different donors, identified as donor 4, donor 72 and donor 234, were added at 100,000 cells/well. The co-cultures were incubated at 37° C. for 16 hours. After incubation, the plates were removed from the incubator and allowed to equilibrate to room temperature over 30 minutes. Bio-Glo luciferase substrate (Promega) was added to each well and the mixture was equilibrated for 10 minutes at room temperature protected from light. Luminescence or relative light units (RFU) was quantified on an EnVision multi-channel reader (Perkin Elmer) with an ultrasensitive luminescent detector. Percent specific kill was calculated as [(About Treatment ECE - Media Only ECE)/Media Only ECE] x 100.

results

In FIG. 99A and B depict the effect of treatment with anti-TIGIT antibody and CHA.7.518.1.H4 (S241P) on killing Mel624 and Raps05.04 cells, respectively. When added separately to co-culture, both the anti-TIGIT antibody and CHA.7.518.1.H4 (S241P) induced significant T-cell killing of tumor cell lines compared to the isotype control antibody. For antiTIGIT antibodies, the percentage specific kill ranged from 19% to 41% for Mel624 cells and from 3% to 44% for Raps05.04 cells between the three different CMV-reactive donors tested. For CHA.7.518.1.H4 (S241P), the specific killing ranged from 16 to 20% for Mel624 cells and from 0.21 to 29% for Raps05.04 cells. In some cases, an additive effect on tumor cell killing was observed with combination therapy with an anti-TIGIT antibody and CHA7.518.1.H4(S241P).

To determine if the effect of anti-TIGIT antibody and CHA.7.518.1.H4 (S241P) on killing tumor cells was dose-dependent, analysis was performed using 10-point, 2-fold dilution series for each antibody, starting from 0, 5 μg/ml for anti-TIGIT antibodies, and 10 μg/ml for CHA.7.518.1.H4 (S241P) (Fig. 100). The death of Mel624 was reduced in a dose dependent manner when either the anti-TIGIT antibody, BM26 or CPA.9.086, was combined with CHA.7.518.1.H4 (S241P). A stronger kill was observed for the combination of CPA.9.086 and CHA.7.518.1.H4(S241P) with EC ₅₀ 0.40 ± 0.49 nM compared to the combination of BM26 and CHA.7.518.1.H4(S241P) with EC ₅₀ 2.6 ± 1.7 nM.

T. Example 20: Biophysical measurement of KD

KinExA equilibrium experiments (kinetic exclusion assay) were performed using a KinExA 3200 instrument (Sapidyne Instruments, Boise, Idaho, USA) at 22°C. Recombinant His-labeled human TIGIT was obtained from Sino Biologicals (Beijing, China) and reconstituted in 1XPBS. All antigen and antibody samples for KinExA assay were prepared in degassed PBST buffer (PBS with 0.05% Tween 20) with 100 μg/ml filtered BSA and 0.02% sodium azide. The secondary detection antibody used was Alexa Flour 647 labeled goat anti-human IgG (H + L) (Jackson ImmunoResearch Laboratories) diluted 400-700-fold in PBST buffer (with BSA and azide) described above from 0.5 mg/ml in IX PBS, pH 7.4. For each KinExA experiment, ~20 µg of human TIGIT was diluted in 1 ml of 50 mM sodium carbonate, pH 9.2, which was added directly to 50 mg of azlactone beads (Ultralink Support, Thermo Scientific, Rockford, IL, USA) and stirred overnight at 4 °C. After mixing, the beads were washed once with 1 M Trisbuffer, pH 8.5, containing 10 mg/ml BSA and moved for one hour at room temperature in the same buffer. The cross-linked beads were added to the bead cuvette in the KinExA instrument and diluted to ~30 ml IX HBS-N (0.01 M Hepes, 0.15 M NaCl, GE Healthcare) containing 0.02% sodium azide, which was also running buffer for the KinExA device. All antigen-bound beads were used immediately after preparation.

For two repeated Kd measurements for CPA.9.086 (Table 1), 14 TIGIT concentrations ranging from 957 aM to 212 pM were equilibrated at room temperature for 72 h with 2.5 pM CPA.9.086 binding sites and 1.8 pM CPA.9.086 binding sites. For CPA.9.083, 14 TIGIT concentrations ranging from 478 aM to 196 pM were equilibrated within ~72 h with 1.8 pM CPA.9.083 binding sites. For repeated measurements of the BM26 hIgG4 reference antibody, 14 TIGIT concentrations ranging from 9.6 pM to 3.53 nM were equilibrated over ~72 h with 20 pM BM26 binding sites and 8.0 pM BM26 binding sites. For CHA.9.547.13, 14 TIGIT concentrations ranging from 10.5 fM to 2.2 nM were equilibrated within ~72 h with 8 pM of MAb CHA.9.547.13 binding sites. The volume flowing through the bag of beads for each equilibrated sample for

-

Claims (1)

of all experiments, ranged from 4 to 11 ml at a flow rate of 0.25 ml/min. The data were consistent with a 1:1 standard equilibrium binding model using KinExA Pro software (version 4.2.10, Sapidyne Instruments) to estimate Kd and obtain a 95% confidence interval (CI) of the smoothed curve. results Both CPA.9.083 and CPA.9.086 are associated with human TIGIT with femtomolar binding affinity, while CHA.9.547.13 and BM26 are associated with picomolar affinity. Thus, CPA.9.083 and CPA.9.086 associated with human TIGIT with the highest affinity of the four different antibodies tested. Table 1 KD measurements of human anti-TIGIT hIgG4 antibodies determined by KinExA Antibody K _D ± 95% CI (n = 1) K _D ± 95% CI (n = 2) SNA.9.547.13 18.8 ± 5.8 pM Undefined CPA.9.083 694 ± 277 fM Undefined CPA.9.086 553 ± 230 fM 665 ± 378 fM VM26 8.2 ± 2.8 pM 11.2±3.6pM U. Example 21: Development and Performance Characterization of CPA.9.086, a Novel Therapeutic Antibody Targeting the TIGIT Immune Checkpoint BACKGROUND OF THE INVENTION: TIGIT is a co-inhibitory receptor that is highly expressed on lymphocytes, including effector and regulatory CD4+ T cells (Treg), effector CD8+ T cells, and NK cells that infiltrate various types of tumors. Involvement of TIGIT with its known ligands, poliovirus receptors (PVR) and PVR-like proteins (PVRL2 and PVRL3) directly inhibits lymphocyte activation. PVR is also widely expressed in tumors, indicating that the TIGIT-PVR signaling axis may be the dominant immunologic surveillance evasion mechanism for cancer. We report herein the biophysical and functional characterization of CPA.9.086, a therapeutic antibody targeting TIGIT. We also demonstrate that co-blockade of TIGIT and the novel checkpoint inhibitor PVRIG increases T cell responses. Materials and Methods: Human phage display and mouse hybridoma antibody technology campaigns were conducted to identify therapeutic anti-TIGIT antibodies. The resulting antibodies were evaluated for their ability to bind to recombinant and cell surface expressed human TIGIT with high affinity. The cross-reactivity of cynomolgus macaque and mouse TIGIT antibodies was also investigated. Subpopulations of antibodies that bind with high affinity to human TIGIT and have cross-reactivity with cynomolgus TIGIT were further characterized by their ability to block the interaction between TIGIT and PVR. Blocking antibodies were screened for their ability to increase antigen-specific T cell activation in vitro either alone or in combination with an anti-PVRIG antibody, CHA.7.518.1.H4(S241P). Results: A major antibody, CPA.9.086, was identified that binds to human TIGIT with high femtomolar affinity. This TIGIT-related antibody is endogenously expressed on human CD8+ T cells with higher affinity than the reference antibodies tested, and was also cross-reactive to both cynomolgus and mouse TIGIT. When tested for in vitro activity, CPA.9.086 increased cytokine secretion and tumor cell killing by CMV-specific CD8+ T cells with superior or equivalent activity to the tested reference antibodies. The combination of CPA.9.086 with anti-PD1 antibody or CHA.7.518.1.H4(S241P) resulted in increased CMV-specific activity of CD8+ T cells. In addition, we have demonstrated that TIGIT is predominantly expressed on Tregs and effector CD8+ T cells from solid tumors compared to peripheral blood, indicating that these populations are likely to be preferentially targeted by CPA.9.086. Conclusion: The development of a very high affinity TIGIT antagonist antibody, CPA.9.086, which is currently in preclinical development, has been described. We postulate that the femtomolar affinity of CPA.9.086 may lead to lower and lower dosing frequency in patients. CPA.9.086 can enhance human T cell activation either alone or in combination with other anti-checkpoint antibodies. Thus, these data demonstrate the utility of targeting TIGIT, PD1, and PVRIG for cancer treatment. CLAIM 1. Composition for the treatment of cancer by activating T cells and/or NK cells, containing an antigen-binding domain that binds to human TIGIT (SEQ ID NO: 97), containing: a) a heavy chain variable domain comprising SEQ ID NO: 160; And b) a light chain variable domain comprising SEQ ID NO: 165, wherein the composition further comprises a pharmaceutically acceptable carrier. -