[0001] This application claims priority to US Application Serial No. 62/376,334, filed on August 17, 2016, US Application Serial No. 62/513,771 filed on June 1, 2017, US
Application Serial No. 62/376,335, filed on August 17, 2016, US Application Serial No. 62/417,217, filed on November 3, 2016, US Application Serial No. 62/513,775, filed on June 1, 2017, US Application Serial No. 62/477,974, filed on March 28, 2017, US Application Serial No. 62/513,916, filed on June 1, 2017, and US Application Serial No. 62/538,561, filed on July 28, 2017, all of which are incorporated by reference herein in their entireties.

SEQUENCE LISTING

[0002] The instant application contains a Sequence Listing which has been submitted electronically in ASCII format and is hereby incorporated by reference in its entirety. Said ASCII copy, created on August 17, 2017, is named 114386-5008-WO_SL.txt and is 590,436 bytes in size.

II. BACKGROUND OF THE INVENTION

[0003] Naive T cells must receive two independent signals from antigen-presenting cells (APC) in order to become productively activated. The first, Signal 1, is antigen-specific and occurs when T cell antigen receptors encounter the appropriate antigen-MHC complex on the APC. The fate of the immune response is determined by a second, antigen-independent signal (Signal 2) which is delivered through a T cell costimulatory molecule that engages its APC-expressed ligand. This second signal could be either stimulatory (positive costimulation) or inhibitory (negative costimulation or coinhibition). In the absence of a costimulatory signal, or in the presence of a coinhibitory signal, T-cell activation is impaired or aborted, which may lead to a state of antigen-specific unresponsiveness (known as T-cell anergy), or may result in T-cell apoptotic death.

[0004] Costimulatory molecule pairs usually consist of ligands expressed on APCs and their cognate receptors expressed on T cells. The prototype ligand/receptor pairs of costimulatory molecules are B7/CD28 and CD40/CD40L. The B7 family consists of structurally related, cell-surface protein ligands, which may provide stimulatory or inhibitory input to an immune response. Members of the

B7 family are structurally related, with the extracellular domain containing at least one variable or constant immunoglobulin domain.

[0005] Both positive and negative costimulatory signals play critical roles in the regulation of cell-mediated immune responses, and molecules that mediate these signals have proven to be effective targets for immunomodulation. Based on this knowledge, several therapeutic approaches that involve targeting of costimulatory molecules have been developed, and were shown to be useful for prevention and treatment of cancer by turning on, or preventing the turning off, of immune responses in cancer patients and for prevention and treatment of autoimmune diseases and inflammatory diseases, as well as rejection of allogenic transplantation, each by turning off uncontrolled immune responses, or by induction of "off signal" by negative costimulation (or coinhibition) in subjects with these pathological conditions.

[0006] Manipulation of the signals delivered by B7 ligands has shown potential in the treatment of autoimmunity, inflammatory diseases, and transplant rejection. Therapeutic strategies include blocking of costimulation using monoclonal antibodies to the ligand or to the receptor of a costimulatory pair, or using soluble fusion proteins composed of the costimulatory receptor that may bind and block its appropriate ligand. Another approach is induction of co-inhibition using soluble fusion protein of an inhibitory ligand. These approaches rely, at least partially, on the eventual deletion of auto- or allo-reactive T cells (which are responsible for the pathogenic processes in autoimmune diseases or transplantation, respectively), presumably because in the absence of costimulation (which induces cell survival genes) T cells become highly susceptible to induction of apoptosis. Thus, novel agents that are capable of modulating costimulatory signals, without compromising the immune system's ability to defend against pathogens, are highly advantageous for treatment and prevention of such pathological conditions.

[0007] Costimulatory pathways play an important role in tumor development. Interestingly, tumors have been shown to evade immune destruction by impeding T cell activation through inhibition of costimulatory factors in the B7-CD28 and TNF families, as well as by attracting regulatory T cells, which inhibit anti-tumor 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. Rev. 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. Cane. Res. 13(13): 4016-4025). Such tumor expressed co-stimulatory molecules have become attractive cancer biomarkers and may serve as tumor-associated antigens (TAAs). Furthermore, costimulatory pathways have been identified as immunologic checkpoints that attenuate T cell dependent immune responses, both at the level of

initiation and effector function within tumor metastases. As engineered cancer vaccines continue to improve, it is becoming clear that such immunologic checkpoints are a major barrier to the vaccines' ability to induce therapeutic anti-tumor responses. In that regard, costimulatory molecules can serve as adjuvants for active (vaccination) and passive (antibody -mediated) cancer immunotherapy, providing strategies to thwart immune tolerance and stimulate the immune system.

[0008] Over the past decade, agonists and/or antagonists to various costimulatory proteins have been developed for treating autoimmune diseases, graft rejection, allergy and cancer. For example, CTLA4-Ig (Abatacept, Orencia®) is approved for treatment of RA, mutated CTLA4-Ig (Belatacept, Nulojix®) for prevention of acute kidney transplant rejection and by the anti-CTLA4 antibody (Ipilimumab, Yervoy®), recently approved for the treatment of melanoma. Other costimulation regulators have been approved, such as the anti-PD-1 antibodies of Merck (Keytruda®) and BMS (Opdivo®), have been approved for cancer treatments and are in testing for viral infections as well.

[0009] However, while monotherapy with anti-checkpoint inhibitor antibodies have shown promise, 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)) examining tumor-infiltrating lymphocytes (TILs) have shown that TILs commonly 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 with monospecific full-length antibodies is likely nondiscriminatory with regards to de-repression of tumor-reactive TILs versus autoantigen-reactive single expressing T cells that are assumed to contribute to autoimmune toxicities.

[0010] One target of interest is PVRIG. PVRIG, also called Poliovirus Receptor Related

Immunoglobulin Domain Containing Protein, Q6DKI7 or C7orfl5, is a transmembrane domain protein of 326 amino acids in length, with a signal peptide (spanning from amino acid 1 to 40), an extracellular domain (spanning from amino acid 41 to 171), a transmembrane domain (spanning from amino acid 172 to 190) and a cytoplasmic domain (spanning from amino acid 191 to 326). PVRIG binds to Poliovirus receptor-related 2 protein (PVLR2, also known as nectin-2, CD 112 or herpesvirus entry mediator B, (HVEB) a human plasma membrane glycoprotein), the binding partner of PVRIG.

[0011] Another target of interest is TIGIT. TIGIT is a coinhibitory receptor that is highly expressed on effector & regulatory (Treg) CD4+ T cells, effector CD8+ T cells, and NK cells. TIGIT has been shown to attenuate immune response by (1) direct signaling, (2) inducing ligand signaling, and (3) competition with and disruption of signaling by the costimulatory receptor CD226 (also known as DNAM-1). TIGIT signaling has been the most well-studied in NK cells, where it has been demonstrated that engagement with its cognate ligand, poliovirus receptor (PVR, also known as

CD 155) directly suppresses NK cell cytotoxicity through its cytoplasmic ITIM domain. Knockout of the TIGIT gene or antibody blockade of the TI GIT/PVR interaction has shown to enhance NK cell killing in vitro, as well as to exacerbate autoimmune diseases in vivo. In addition to its direct effects on T- and NK cells, TIGIT can induce PVR-mediated signaling in dendritic or tumor cells, leading to the increase in production of anti-inflammatory cytokines such as IL10. In T-cells TIGIT can also inhibit lymphocyte responses by disrupting homodimerization of the costimulatory receptor CD226, and by competing with it for binding to PVR.

[0012] TIGIT is highly expressed on lymphocytes, including Tumor Infiltrating Lymphocytes (TILs) and Tregs, that infiltrate different types of tumors. PVR is also broadly expressed in tumors, suggesting that the TIGIT -PVR signaling axis may be a dominant immune escape mechanism for cancer. Notably, TIGIT expression is tightly correlated with the expression of another important coinhibitory receptor, PD1. TIGIT and PD1 are co-expressed on the TILs of numerous human and murine tumors. Unlike TIGIT and CTLA4, PD1 inhibition of T cell responses does not involve competition for ligand binding with a costimulatory receptor.

[0013] Accordingly, TIGIT is an attractive target for monoclonal antibody therapy, and in addition in combination with additional antibodies including anti-PVRIG antibodies.

III. BRIEF SUMMARY OF THE INVENTION

[0014] Accordingly, in one aspect, the invention provides compositions comprising an antigen binding domain that binds to human TIGIT (SEQ ID NO:97) comprising a variable heavy domain comprising SEQ ID NO: 160 and a variable light domain comprising SEQ ID NO: 165. Additionally, the antigen binding domain comprises a variable heavy domain comprising SEQ ID NO: 150 and a variable light domain comprising SEQ ID NO: 155. Additionally, the antigen binding domain comprises a variable heavy domain comprising SEQ ID NO: 560 and a variable light domain comprising SEQ ID NO:565.

[0015] In a further aspect, the invention provides composition comprising antibodies comprising a heavy chain comprising VH-CHl-hinge-CH2-CH3, wherein said VH comprises SEQ ID NO: 160 and a light chain comprising VL-VC, wherein said VL comprising SEQ ID NO: 165 and VC is either kappa or lambda. Additionally, the antibody can comprise a heavy chain comprising VH-CHl-hinge-CH2-CH3, wherein the VH comprises SEQ ID NO: 150; and a light chain comprising VL-VC, wherein said VL comprising SEQ ID NO: 159 and VC is either kappa or lambda. Additionally, the antibody can comprise a heavy chain comprising VH-CHl-hinge-CH2-CH3, wherein the VH comprises SEQ ID NO:560; and a light chain comprising VL-VC, wherein said VL comprising SEQ ID NO: 565 and VC is either kappa or lambda.

[0016] In some aspects, the sequence of the CHl-hinge-CH2-CH3 is selected from human IgGl, IgG2 and IgG4, and variants thereof. In some aspects, the heavy chain has SEQ ID NO: 164 and the light chain has SEQ ID NO: 169.

[0017] In an additional aspect, the compositions can further comprise a second antibody that binds to a human checkpoint receptor protein, which can be human PD-1 or human PVRIG. The second antibody can comprises an antigen binding domain comprising a variable heavy domain comprising SEQ ID NO:5 and a variable light domain comprising SEQ ID NO: 10, or a heavy chain having SEQ ID NO:9 and a light chain having SEQ ID NO: 14.

[0018] In a further aspect, the invention provides nucleic acid compositions comprising a first nucleic acid encoding a variable heavy domain comprising SEQ ID NO: 160 and a second nucleic acid encoding a variable light domain comprising SEQ ID NO: 165. Alternatively, the nucleic acid compositions comprise a first nucleic acid encoding a variable heavy domain comprising SEQ ID NO: 150 and a second nucleic acid encoding a variable light domain comprising SEQ ID NO: 155. Alternatively, the nucleic acid compositions comprise a first nucleic acid encoding a variable heavy domain comprising SEQ ID NO:560 and a second nucleic acid encoding a variable light domain comprising SEQ ID NO:565.

[0019] In a further aspect, the invention provides expression vector compositions comprising these nucleic acid compositions are provided as well, such as a first expression vector comprising a first nucleic acid and a second expression vector comprising a second nucleic acid, or alternatively an expression vector that comprises both first and second nucleci acids.

[0020] In an additional aspect, the invention provides host cells comprising the expression vector compositions, and methods of making the antibodies comprising culturing the host cells under conditions wherein the antibodies are produced and recovering the antibody.

[0021] In a further aspect the invention provides anti-PVRIG antibodies comprising a heavy chain having SEQ ID NO:9 and a light chain having SEQ ID NO: 14. The invention further provides antibodies having a heavy chain having SEQ ID NO: 19; and a light chain having SEQ ID NO:24.

[0022] In an additional aspect, an anti-PVRIG antibody (either CHA.7.518.1.H4(S241P) or CHA.7.538.1.2.H4(S241P) are co-administered with a second antibody that binds to a human checkpoint receptor protein, such as an antibody that binds PD-1.

[0023] In a further aspect, an anti-PVRIG antibody (either CHA.7.518.1.H4(S241P) or CHA.7.538.1.2.H4(S241P)) are co-administered with a second antibody that binds to a human checkpoint receptor protein, such as an antibody that binds human TIGIT, such as CPA.9.086 or CPA.9.083 or CHA.9.547.13.

[0024] In a further aspect, the invention provides nucleic acid compositions comprising a first nucleic acid encoding the heavy chain of either CHA.7.518.1.H4(S241P) or CHA.7.538.1.2.H4(S241P)) and a second nucleic acid encoding the light chain of either CHA.7.518.1.H4(S241P) or

CHA.7.538.1.2.H4(S241P), respectively.

[0025] In a further aspect, the invention provides expression vector compositions comprising these nucleic acid compositions are provided as well, such as a first expression vector comprising a first nucleic acid and a second expression vector comprising a second nucleic acid, or alternatively an expression vector that comprises both first and second nucleci acids.

[0026] In an additional aspect, the invention provides host cells comprising the expression vector compositions, and methods of making the antibodies comprising culturing the host cells under conditions wherein the antibodies are produced and recovering the antibody.

[0027] In a further aspect, the invention provides methods comprising: a) providing a cell population from a tumor sample from 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) an isotype control; c) running fluorescence 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; wherein if the percentage of positive cells is > 1% for all 4 receptors, e) administering antibodies to TIGIT and PD-1 to said patient.

[0028] In an additional aspect, the invention provides methods comprising: a) providing a cell population from a tumor sample from 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) an isotype control; c) running fluorescence 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; wherein if the percentage of positive cells is > 1% for all 4 receptors, e) administering antibodies to PVRIG and PD-1 to said patient.

[0029] In a further aspect, the invention provides methods comprising a) providing a cell population from a tumor sample from 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) an isotype control; c) running fluorescence 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; wherein if the percentage of positive cells is > 1% for all 4 receptors, e) administering antibodies to PVRIG and TIGIT to said patient.

IV. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 depicts the full-length sequence of human PVRIG (showing two different methionine starting points). The signal peptide is underlined, the ECD is double underlined.

[0031] Figure 2 depicts the sequence of the human Poliovirus receptor-related 2 protein (PVLR2, also known as nectin-2, CD 112 or herpesvirus entry mediator B, (HVEB)), the binding partner of PVRIG. PVLR2 is a human plasma membrane glycoprotein.

[0032] Figure 3A and B depicts the variable heavy and light chains as well as the vhCDRl, vhCDR2, vhCDR3, vlCDRl, vlCDR2 and vlCDR3 sequences of each of the enumerated CHA antibodies of the invention, CHA.7.518.1.H4(S241P), and CHA.7.538.1.2.H4(S241P).

[0033] Figure 4A and B PVRIG antibodies increase T cell proliferation in the MLR. The percentages of CFSE low cells are shown from MLR assays treated with the indicated PVRIG antibodies. Each graph represents one individual CD3 T cell donor. The experiments are described in Example 23 of USSN 15/048,967, incorporated by reference herein.

[0034] Figure 5 A and B PVRIG hybridoma antibody binding characteristics to HEK hPVRIG engineered cell lines, HEK parental cells, and Jurkat cells. HEK OE denotes HEK hPVRIG cells, HEK par denotes HEK parental cells. For Jurkat data, gMFIr indicates the fold difference in geometric MFI of PVRIG antibody staining relative to their controls. Concentration indicates that at which the gMFIr was calculated. No binding indicates antibody does not bind to the tested cell line. Highlighted antibodies are the 'top four' antibodies of interest.

[0035] Figure 6A and B PVRIG hybridoma antibody binding characteristics to primary human PBMC, cyno over-expressing cells, and cyno primary PBMC. Expi cyno OE denotes expi cells transiently transfected with cPVRIG, expi par denotes expi parental cells. gMFIr indicates the fold difference in geometric MFI of PVRIG antibody staining relative to their controls. Concentrations indicate that at which the gMFIr was calculated. Not tested indicates antibodies that were not tested due to an absence of binding to human HEK hPVRIG, expi cPVRIG cells, or not meeting binding requirements to PBMC subsets. Highlighted antibodies are the 'top four' antibodies of interest. The experiments are described in Example 21 of USSN 15/048,967, incorporated by reference herein.

[0036] Figure 7A and B Summary of blocking capacity of PVRIG antibodies in the FACS-based competition assay. The IC50 of inhibition is indicated. No IC50 indicates that these antibodies are non-blockers. Highlighted antibodies are the 'top four' antibodies of interest. The experiments are described in Example 21 of USSN 15/048,967, incorporated by reference herein.

[0037] Figure 8A and B TILs were co-cultured with melanoma cells 624 at 1: 1 E:T for 18hr in the presence of anti-PVRIG Ab (CPA.7.021; lOug/ml) , anti-TIGIT (10A7 clone; lOug/ml) or in combination. Supernatant was collected and tested in Thl Th2 Thl7 cytometric bead array assay to detect secreted cytokines. IFNy (A) and TNF (B) levels were detected. Treatments were compared by Student's t-test (*P < 0.05, **P < 0.01) of triplicate samples.

[0038] Figure 9A to F MART-1 or 209 TILs were co-cultured with melanoma cells 624 at 1 : 1 E:T for 18hr in the presence of anti-PVRIG Ab (CPA.7.021; lOug/ml) , anti-DNAMl (DX11 clone, BD Biosciences Cat. No. 559787; lOug/ml) or in in combination. Supernatant was collected and tested in Thl Th2 Thl7 cytometric bead array assay to detect secreted cytokines. ΙΚΝγ (A,D) and TNF (B,E) levels were detected. TILs were stained for surface expression of CD137 (C,F).

[0039] Figure 10A and B TILs (F4) were co-cultured with melanoma cells 624 at 1 :3 E:T for 18hr in the presence of anti-PVRIG Ab (CPA.7.021; lOug/ml) , anti-TIGIT (10A7 clone; lOug/ml), anti-PDl (mAb 1B8, Merck; lOug/ml) or in combination. Supernatant was collected and tested in Thl Th2 Thl7 cytometric bead array assay to detect secreted cytokines. IFNy (A) and TNF (B) levels were detected.

[0040] Figure 11 A to E depict four humanized sequences for each of CHA.7.518, CHA.7.524, CHA.7.530, CHA.7.538 1 and CHA.7.538 2. All humanized antibodies comprise the H4(S241P) substitution. Note that the light chain for CHA.7.538 2 is the same as for CHA.7.538 1. The "HI" of each is a "CDR swap" with no changes to the human framework. Subsequent sequences alter framework changes shown in larger bold font. CDR sequences are noted in bold. CDR definitions are AbM from website www .biotnf.or8.-tik/abs,''. Human germline and joining sequences from IMGT® the international ImMunoGeneTics® information system www.imgt.org (founder and director: Marie-Paule Lefranc, Montpellier, France). Residue numbering shown as sequential (seq) or according to Chothia from website www.bioinf.org.uk/abs/ (AbM). "b" notes buried sidechain; "p" notes partially buried; "i" notes sidechain at interface between VH and VL domains. Sequence differences between human and murine germlines noted by asterisk (*). Potential additional mutations in frameworks are noted below sequence. Potential changes in CDR sequences noted below each CDR sequence as noted on the figure (# deamidation substitutions: Q/S/A; these may prevent asparagine (N) deamidation. @ tryptophan oxidation substitutions: Y/F/H; these may prevent tryptophan oxidation; @ methionine oxidation substitutions: L/F/A).

[0041] Figure 12A to E depicts a collation of the humanized sequences of three CHA antibodies: CHA.7.518, CHA.7.538.1, and CHA.7.538.2.

[0042] Figure 13. depicts schemes for combining the humanized VH and VL CHA antibodies. The "chimVH" and "chimVL" are the mouse variable heavy and light sequences attached to a human IgG constant domain.

[0043] Figure 14. PVRIG hybridoma antibody binding characteristics to primary human PBMC, cyno over-expressing cells, and cyno primary PBMC. Expi cyno OE denotes expi cells transiently transfected with cPVRIG, expi par denotes expi parental cells. gMFIr indicates the fold difference in geometric MFI of PVRIG antibody staining relative to their controls. Concentrations indicate that at which the gMFIr was calculated. Not tested indicates antibodies that were not tested due to an absence of binding to human HEK hPVRIG, expi cPVRIG cells, or not meeting binding requirements to PBMC subsets. Highlighted antibodies are four antibodies for which humanization was done (See Figure 24). The experiments are described in Example 21 of USSN 15/048,967, incorporated by reference herein.

[0044] Figure 15. Summary of blocking capacity of PVRIG antibodies in the FACS-based competition assay. The IC50 of inhibition is indicated. No IC50 indicates that these antibodies are non-blockers. Highlighted antibodies are four antibodies for which humanization was done (See Figure 24).

[0045] Figure 16. Summary of the activity of select PVRIG antibodies in NK cell cytotoxicity assays against Reh and MOLM-13 cells. Fold change in cytotoxicity relative to control was calculated by dividing the absolute level of killing (%) in the condition with PVRIG antibody, by the absolute level of killing (%) with control antibody. Fold change is calculated from the 5: 1 effector to target ratio.

[0046] Figure 17. Sequence alignment of PVRIG orthologs. Aligned sequences of the human, cynomolgus, marmoset, and rhesus PVRIG extra-cellular domain. The differences between human and cynomolgus are highlighted in yellow.

[0047] Figure 18. Binding of anti-human PVRIG antibodies to cyno, human, cyno/human hybrid PVRIG variants. Binding of antibodies to wild type cyno PVRIG (·), H61R cyno PVRIG (■), P67S cyno PVRIG ( ), L95R/T97I cyno PVRIG (τ), and wild type human PVRIG (♦) are shown. The ELISA signals are plotted as a function of antibody concentration.

[0048] Figure 19. Correlation of epitope group and cyno cross-reactivity of anti-human PVRIG antibodies.

[0049] Figure 20A to B (A) Specificity of CHA.7.518.1.H4(S241P) towards HEK cells engineered to overexpress PVRIG and HEK parental cells. Data shows absolute geometric MFI (gMFI)

measurements as a function of increasing antibody concentration. (B) Specificity of

CHA.7.538.1.2.H4(S241P) towards HEK cells engineered to overexpress PVRIG and HEK parental cells. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration.

[0050] Figure 21 A to B illustrates 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. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration. Isotype staining is shown as a negative control. (B) Binding of CHA.7.538.1.2.H4(S241P) to Jurkat cells. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration. Isotype staining is shown as a negative control. Both antibodies are able to bind Jurkat cells with a comparable affinity to HEK hPVRIG cells.

[0051] Figure 22 illustrates the ability of CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P) to bind CD8 T cells that were expanded by exposure to CMV peptide (494-503, NLVPMVATV) and 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 CD 8 T cells. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration. Isotype staining is shown as a negative control.

[0052] Figure 23A to B. (A) Specificity of CHA.7.518.1.H4(S241P) towards expi cells engineered to overexpress cynomolgus PVRIG and expi parental cells. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration. Specificity of

CHA.7.538.1.2.H4(S241P) towards expi cells engineered to overexpress cynomolgus PVRIG and expi parental cells. Data shows absolute geometric MFI (gMFI) measurements as a function of increasing antibody concentration.

[0053] Figure 24A to B. (A) Blocking of PVRIG Fc to HEK cells by CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P). Data shows the percentage of PVRIG Fc binding to HEK cells as a function of increasing antibody concentration relative to maximum PVRIG Fc-induced signal and secondary only background. (B) Effect of CHA.7.544 on the binding of PVRIG Fc to HEK cells. Data shows the absolute gMFI derived from PVRIG Fc binding to HEK cells in the presence of escalating concentrations of CHA.7.544. The amount of PVRIG Fc binding was detected by an anti-mouse Fc secondary conjugated to Alexa 647.

[0054] Figure 25 A to B. (A) Blocking of PVRL2 Fc to HEK hPVRIG cells by

CHA.7.518.1.H4(S241P), CHA.7.538.1.2.H4(S241P), and CHA.7.530.3. Data shows the percentage of PVRL2 Fc binding to HEK hPVRIG cells as a function of increasing antibody concentration

relative to maximum PVRL2 Fc-induced signal and secondary only background. (B) Effect of CHA.7.544 on PVRL2 Fc binding to HEK hPVRIG cells. Data shows the percentage of PVRL2 Fc binding to HEK hPVRIG cells as a function of increasing antibody concentration relative to maximum PVRL2 Fc-induced signal and secondary only background.

[0055] Figure 26. Shows the percentage of Alexa 647 conjugated CHA.7.518. l.H4(S241P) and CHA.7.538.1.2.H4(S241P) binding relative to their maximum signal upon pre-incubation of Jurkat cells with unconjugated CHA.7.518.1.H4(S241P), CHA.7.538.1.2.H4(S241P) and an isotype control.

[0056] Figure 27A to B A) Humanized PVRIG antibodies, CHA.7.518. l.H4(S241P) and

CHA.7.538.1.2.H4(S241P), increase CD4+ T cell proliferation. Representative data (n>2) shows the percentage of CFSE low, proliferating CD4+ T cells (mean plus standard deviation) from a single human CD4+ T cell donor when co-cultured with the CHO-S OKT3 hPVRL2 cells in the presence of an anti-DNAM-1 antibody or different anti-PVRIG antibodies or IgG isotype controls. The dashed line indicates the baseline percentage of CFSE low, CD4+ T cells proliferating after treatment with the human IgG4 isotype control antibody. The numbers refer to the percent increase or decrease in proliferation of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies (B) Humanized PVRIG antibodies, CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P), increase CD4+ T cell proliferation in an hPVRL2-dependent manner. Representative data (n>2) shows the percentage of CFSE low, proliferating CD4+ T cells (mean plus standard deviation) from a single human CD4+ T cell donor in response to co-culture with the CHO-S OKT3 parental, or CHO-S OKT3 hPVRL2 cells in the presence of an anti-DNAM-1 antibody or different anti-PVRIG antibodies or IgG isotype controls. The dashed line indicates the baseline percentage of CFSE low CD4+ T cells proliferating after treatment with either the human IgG4 or the mouse IgGl isotype antibodies. The numbers refer to the percent increase or decrease in proliferation of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies.

[0057] Figure 28A to C. (A) Humanized PVRIG antibodies, CHA.7.518.1.H4(S241P) and

CHA.7.538.1.2.H4(S241P), increase CD8+ T cell proliferation. Representative data (n>2) shows the percentage of CFSE low, proliferating CD8+ T cells (mean plus standard deviation) from a single human CD8+ T cell donor (Donor 232) when co-cultured with the CHO-S OKT3 hPVRL2 cells in the presence of an anti-DNAM-1 antibody or different anti-PVRIG antibodies or IgG isotype controls. The dashed line indicates the baseline percentage of CFSE low, CD8+ T cells proliferating after treatment with the mouse IgGl or human IgG4 isotype antibodies. The numbers refer to the percent increase or decrease in proliferation of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies. (B) Humanized PVRIG antibodies, CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P), increase CD8+ T cell proliferation.

Representative data (n>2) shows the percentage of CFSE low, proliferating CD8+ T cells (mean plus standard deviation) from a single human CD8+ T cell donor (Donor 234) when co-cultured with the CHO-S OKT3 hPVRL2 cells in the presence of an anti-DNAM-1 antibody or different anti-PVRIG antibodies or IgG isotype controls. The dashed line indicates the baseline percentage of CFSE low, CD8+ T cells proliferating after treatment with the mouse IgGl or human IgG4 isotype antibodies. The numbers refer to the percent increase or decrease in proliferation of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies. (C) Humanized PVRIG antibodies, CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P), increase ΙΚΝγ secretion from CD8+ T cells. Representative data (n>2) shows the pg/ml of ΙΚΝγ produced (mean plus standard deviation) by three different human CD8+ T cell donors (Donors 231, 232, and 234) when co-cultured with the CHO-S OKT3 hPVRL2 cells in the presence of an anti-DNAM-1 antibody or different anti-PVRIG antibodies or IgG isotype controls. The dashed line indicates the baseline IFNy production following treatment with the human IgG4 isotype antibody. The numbers refer to the percent increase in IFNy secretion of the anti-PVRIG antibody treatments compared to the IgG4 isotype control.

[0058] Figure 29. Humanized PVRIG antibodies, CHA.7.518.1.H4(S241P) and

CHA.7.538.1.2.H4(S241P), consistently increase CD4+ T cell proliferation across multiple donors, while CHA.7.530.3 and CHA.7.544 do not. The percent proliferation relative to the isotype control was calculated by dividing the percentage of CFSE low, CD4+ T cells after PVRIG antibody treatment over the isotype antibody treatment for each donor. The percent proliferation for the isotype antibody treatment was set at zero. Each symbol in the graph represents a different donor.

[0059] Figure 30A to D. (A) Dose-dependent effect of the humanized PVRIG antibodies,

CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P), on CD4+ T cell proliferation.

Representative data (n>2) with 2 different human donors shows the mean percentage of proliferating CD4+ T cells following a dose titration of 66nM to 0.726nM with either the human IgG4 isotype, CHA.7.518.1.H4(S241P), or CHA.7.538.1.2.H4(S241P) antibodies. The estimated EC50 is within the single digit nM range. (B) Dose-dependent effect of the humanized PVRIG antibodies,

CHA.7.518.1.H4(S241P) and CHA.7.538.1.2.H4(S241P), on CD8+ T cell proliferation.

Representative data (n>2) with 2 different human donors shows the mean percentage of proliferating CD8+ T cells following a dose titration of 66nM to 0.264nM with either the human IgG4 isotype, CHA.7.518.1.H4(S241P), CHA.7.38.1.2, or CHA.7.544 antibodies. The estimated EC50 is within the single digit nM range.

[0060] Figure 31 A to C. (A) Flow cytometry analysis of TIGIT and PVRIG expression on TILs and PVR, PVRL2 expression on 624 melanoma cell line. Values represent Mean fluorescent intensity (MFI) ratio vs isotype control. (B-C) Representative experiment showing IFNy (B) and TNF (C)

secretion by TILs upon co-cultured with melanoma cells 624 at 1 :3 E:T for 18hr in the presence of isotype control, anti-TIGIT (3(^g/ml) or anti- PVRIG Abs (lOug/ml) as mono treatment (blue histograms) or in combination with anti-TIGIT (green histograms). Percentage of Ab mono treatment effect was compared to isotype control treatment mlgGl and the percentage of Ab combo-treatment effect was compared to anti-TIGIT mono-treatment.

[0061] Figure 32A to H. TILs (209-gpl 00/463 -F4-gp 100) were co-cultured with melanoma cells 624 in 1 :3 E:T for 18hr in the presence of anti PVRIG Abs CHA.7.518.1.H4(S241P) or CHA.7.538 with or without anti-TIGIT (aTIGIT) combo and tested for cytokine secretion. Percentage of Ab treatment effect was compared to isotype control treatment and the mean of 5 experiments (F4) or 6 experiments (209) were plotted. Paired, two tailed T test was calculated for each treatment compared to isotype or in combos- compared to anti-TIGIT alone, p values are indicated.

[0062] Figure 33A to B. (A) Humanized PVRIG antibody, CHA.7.518.1.H4(S241P), and an anti-TIGIT antibody increase CD4+ T cell proliferation compared to single antibody treatments.

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

[0063] Figure 34A to B. (A): The combination of the humanized PVRIG antibody,

CHA.7.518.1.H4(S241P), and the anti-TIGIT antibody increases CD8+ T cell proliferation.

Representative data (n>2) shows the percentage of CFSE low, proliferating CD8+ T cells (mean plus standard deviation) from a representative human CD8+ T cell donor (Donor 232) when co-cultured with the CHO-S OKT3 hPVRL2 cells. The dashed line indicates the baseline percentage of CFSE low, CD8+ T cells proliferating after treatment with the human IgG4 isotype antibody. The numbers refer to the percent increase or decrease in proliferation of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies. (B) The combination of the humanized PVRIG antibody, CHA.7.518.1.H4(S241P), and the anti-TIGIT antibody increases ΙΚΝγ secretion from CD8+ T cells. Representative data (n>2) shows the pg/ml of ΙΚΝγ produced

(mean plus standard deviation) by a representative human CD8+ T cell donor (Donor 232) when co-cultured with the CHO-S OKT3 hPVRL2 cells. The dashed line indicates the baseline IFNy production following treatment with the human IgG4 isotype antibody. The numbers refer to the percent increase or decrease in IFNy secretion of the anti-PVRIG or anti-DNAM-1 antibody treatments, respectively, compared to the relevant isotype control antibodies.

[0064] Figure 35 depicts the design of the experimental system of Example 2(3).

[0065] Figure 36A to C. shows a histogram depicting levels of PVRIG (using Anti-Human PVRIG CHA.7.538.AF647), TIGIT (using Anti-Human TIGIT Cat. 17-9500-41 eBioscience) and DNAM-1 (using Anti-human CD226-APC Cat.338312 biolegend) expression in TILs. Fold of expression is compared to isotype (Iso) control.

[0066] Figure 37. Summarized plot of the effect of anti PVRIG antibodies on the secretion of IFNy from TILs. TILs were co-cultured with CHO-S HLA-A2/B2M cells over-expressing PVRL2 in E:T ratio of 1 :3 for 18hr in the presence of anti PVRIG antibodies (c518, c538 and 544) or with anti TIGIT antibody. Each dot represents an average of data of ΙΚΝγ secretion from the same TIL from different experiments. The percentage indicated is the different between each antibody treatment compared to isotype control. Paired, two tailed T-test was calculated for each treatment compared to 544 or in combos, compared to anti TIGIT alone, p values are indicated. Number of experiments preformed per each TILs; 209(N=3), F4 (N=2), F5(N=3) and MART1(N=2).

[0067] Figure 38. Summarized plot of the effect of c518 and c538 dose response on the secretion of TNF-a from TILs. TILs were co-cultured with CHO-S HLA-A2/B2M cells over-expressing PVRL2 in effector-to-target ratio of 1:3 for 18hr in the presence of anti PVRIG antibodies (c518, c538 or isotype control) as described in Example 2(3).

[0068] Figure 39A to C. TILs were co-cultured with CHO-S HLA-A2/B2M target cells over-expressing PVRL2 in E:T ration of 1:3 for 18hr in the presence of anti PVRIG antibodies (c518, c538 and 544) or with anti TIGIT antibody. The percentage indicated in the above tables is the difference in the effect of cytokine secretion from TILs of each antibody treatment compared to its isotype control. The first experiment is represented in Figure A and B, and the second experiment in Figure C.

[0069] Figure 40. CHO-S OKT3 co-culture assay design. CFSE labeled CD3+ T cells were co-cultured with CHO-S-OKT3-PVRL2 or mock transfected cells for 5d. The effect anti-PVRIG Abs on T cell proliferation and cytokine secretion was analyzed.

[0070] Figure 41. Effect of anti-PVRIG antibodies on IFNy secretion upon CHO-OKT3 PVRL2 cells in responder vs. non-responder donor. CD3+ cells from 2 different donors were co-cultured with CHO-S-PVRL2 cells in 5: 1 E:T for 5d in the presence of anti PVRIG Abs and tested for cytokine secretion and T cells proliferation. (A) 'responder donor' in which we observed an effect to anti PVRIG Abs. (B) 'non-responder donor' in which we do not observed effects to Abs treatment.

[0071] Figure 42. Effect of anti-PVRIG antibodies on CD4 and CD8 proliferation from responder donor. CFSE labeled CD3+ T cells were co-cultured with CHO-S-PVRL2 cells in 5: 1 E:T for 5d in the presence of anti PVRIG Abs or anti-TIGIT Abs. The effect on T cells proliferation gating on CD4 or CD8 was evaluated by flow cytometry. Percentage of proliferating cells (CFSE low) (A) or total cells number (B) of CD4+CFSElow or CD8+CFSE low are presented.

[0072] Figure 43. Shows the effect of anti-PVRIG antibodies on ΙΚΝγ secretion or CD8 proliferation from responder donor. CD3+ cells were co-cultured with CHO-S-PVRL2 cells in 5: 1 E:T for 5d in the presence of anti PVRIG Abs and tested for (A) cytokine secretion and (B) T cells proliferation. Percentage of Ab treatment effect was compared to isotype control treatment and the mean of 5 'responders' donors (responders) is presented. (C) ΙΚΝγ secretion levels from the same 5 donors upon co-culture with CHOS-OKT3 PVRL2 as described in section A and B upon treatment with isotype vs. anti-PVRIG Abs. p value represent ratio paired T test.

[0073] Figure 44 is a summary table of Abs treatment effect across donors tested (n=10).

Percentages indicated represent the effect of Ab treatment on a specific readout (indicated in columns titles) as compared to the relevant isotype control, 'responder' donors (donors #3, 72,226,345 and ES 001) considered as 'responder' which some anti-PVRIG Abs (mainly CHA.7.518) enhanced ΙΚΝγ or proliferation vs. isotype controls.

[0074] Figure 45 A to B depict the results of experiments with several antibodies. The affinities (nM) are shown in A, with the HEK hPVRIG cells being HEK cells transformed with hPVRIG as discussed herein and Jurkat cells expressing endogeneous hPVRIG. (B) depicts the gMFI using 4 different antibodies against Donor 1 primary CD8 T cells and (C) being Donor 2 primary CD8 T cells.

[0075] Figure 46A to B depict interactions of TIGIT with CHO cells. (A) Human TIGIT Fc protein binds to CHO cells. Graded concentrations of human TIGIT Fc and synagis IgGl control were assessed for 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 the scFv of the OKT3 antibody and activated for 5 days. On day 5, CD4 T cells were analysed for expression of PVR and dilution of CFSE.

[0076] Figure 47A to C depict antitumor responses of anti-mP VRlg and anti-PDL-1 antibodies in CT26 tumor model. A-B. Groups of 10 BALB/c mice were subcutaneously injected with 5x 105 CT26

cells. After tumors were measured on day 4, mice were randomized (40 mm3 mean tumor volume per group) and then treated with the designated mAb (100 or 200 ^ig/dose IP) followed by additional doses on days 7, 31, 14, 18 and 21. A. Groups were treated with 6 doses of single agents. Anti-PDL-1 vs control ***p <0.0001. Tumor volumes are represented as the Mean volume + SEM. B, Tumor volumes were measured twice weekly. The number of tumor-free (TF) mice per group is indicated. C. survival proportions of assigned groups; Anti-PDL-1 vs control **p =0.005.

[0077] Figure 48A to C depict antitumor responses of anti-PVRIG and anti-PDL-i antibodies combination in CT26 tumor model. A-B. Groups of 10 BALB/c mice were subcutaneously injected with 5* i05 CT26 cells. After tumors were measured on day 7, mice were randomized (75 mmJ mean tumor volume per group) and then treated with the designated mAb (300 με,/άο$& IP) followed by additional doses on days 11, 14, 18,21 and 25. A. Groups were treated with 6 doses of combined agents. Anti-PDL-l+mAb 407 vs control p = 0.0005; anti-PDL-1 and mAb 406 vs control p=0.056. B. Tumor volumes were measured x3 weekly. The number of tumor-free (TF) mice per group is indicated. C. survival proportions of assigned groups; Anti-PDL-l+mAb 407 vs control *p ==0.0088.

[0078] Figure 49A to D depict the amino acid sequences and the nucleic acid sequence for the variable heavy chain (A and B, respectfully) and the amino acid sequences and the nucleic acid sequence for the variable light chain (C and D, respectfully) for AB-407 (BOJ-5G4-F4).

[0079] Figure 50 depicts the amino acid sequences of the constant domains of human IgGl (with some useful amino acid substitutions), IgG2, IgG3, IgG4, IgG4 with a hinge variant that finds particular use in the present invention, and the constant domains of the kappa and lambda light chains.

[0080] Figure 51 depicts the sequences of human and cynomolgus macaque (referred to as cyno) TIGIT ECD and of the human PVR ECD proteins.

[0081] Figure 52. Shows the flow cytometry binding summary for anti-TIGIT fabs. All unique ELISA positive fabs were analyzed by flow cytometry. The mean fluorescence intensity (MFI) was measured for the human or cyno TIGIT over-expressing Expi293 cells as well as the parental Expi293 cells. The MFI ratio for the target-specific vs off -target binding was calculated. Data for selected clones is shown.

[0082] Figure 53A and B depict the sequences of anti-TIGIT antibodies. Unless otherwise noted, the CDRs utilize the IMGT numbering (including the antibodies of the sequence listing.

[0083] Figure 54. Shows the FACS KD results of anti-TIGIT mAbs binding to Expi293 human TIGIT over-expressing cells as described in Example 12.

[0084] Figure 55. Shows the FACS KD results of mAbs binding to Expi293 cyno TIGIT over-expressing cells.

[0085] Figure 56. Shows the results from Example 14, showing the resulting kinetic rate constants and the equilibrium dissociation constants where data were reliable enough to estimate the binding constants.

[0086] Figure 57A to B show the results of human PVR-Fc variant binding to Expi293 human TIGIT over-expressing cells in Example 4. Figure A (left): Binding curve generated for human PVR-m2aFc construct titrated with Expi293 human TIGT over-expressing cells. The KD and 95% confidence interval are shown. Figure B (right): Binding curve generated for human PVR-hlFc construct titrated with Expi293 human TIGT over-expressing cells. The KD and 95% confidence interval are shown.

[0087] Figure 58. Shows a table of phage antibodies inhibiting human PVR-m2aFc binding to human TIGIT over-expressed on Expi293 cells. mAbs were tested against known blocking (BM26) benchmark antibody, and human IgG4 isotype control (Synagis) antibody. A "Yes" indicates the mAb inhibited hPVR analogous to BM26.

[0088] Figure 59. Shows a table of IC50 values of anti-TIGIT hybridoma antibodies inhibiting binding of human PVR-hlFc to human TIGIT over-expressed on Expi293 cells. Values are representative of one of two independent experiments. The IC50 results for the two independently performed experiments showed a range of only 1.2-2 -fold differences.

[0089] Figure 60. Shows the results of Example 6, that the phage-derived and BM anti-human TIGIT antibodies, CPA.9.027, CPA.9.049, CPA.9.059, BM26, and BM29 increase IL-2 signaling. BM26 and BM29 are both the human IgG4 (hIgG4 with a S241P variant) isotype. Representative data (n>2) shows the RLU (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of Jurkat IL-2-RE luciferase human TIGIT cells and aAPC CHO-K1 human PVR cells. The concentration of each antibody was 10 μg/ml.

[0090] Figure 61. Shows additional results of Example 6, that the phage-derived and BM hIgG4 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 are both the hIgG4 isotype. Representative data (n>2) shows the RLU (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of Jurkat IL-2-RE luciferase human TIGIT cells and aAPC CHO-K1 human PVR cells. A 10 point, 2-fold dilution series starting at 20 μg/ml was used for each antibody.

[0091] Figure 62. Shows the results of Example 6, that the hybridoma-derived and BM 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 are both the mlgGl isotype. The non-blocking anti-human TIGIT antibody, CHA.9.543 does not enhance IL-2 signaling. Representative data (n>2) shows the RLU (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of Jurkat IL-2 -RE luciferase human TIGIT cells and aAPC CHO-KI human PVR cells. The concentration of each antibody was 10 μg/ml.

[0092] Figure 63. Shows the results of Example 6, that the hybridoma-derived and benchmark mlgGl anti-human TIGIT 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 is the mlgGl isotype.

Representative data (n>2) shows the RLU (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of Jurkat IL-2 -RE luciferase human TIGIT cells and aAPC CHO-KI human PVR cells. A 10 point, 2-fold dilution series starting at 20 μg/ml was used for each antibody.

[0093] Figure 64. Shows that the phage, hybridoma and BM anti-human TIGIT antibodies, 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 increase antigen-specific ΙΚΝγ signaling. BM26 is tested as both the hIgG4 and mlgGl isotypes, while BM29 is only tested as the hIgG4 isotype. Representative data (n=2) shows the amount of ΙΚΝγ (mean +/- standard deviation) in the culture supernatant after 24 hour co-culture of CMV-specific CD8+T cells with the Mel624 human PVR cells. The concentration of each antibody was 10 Dg/ml. The Mel624 human PVR used in the assay were pulsed with 0.0033 μg/ml or 0.001 μg/ml peptide.

[0094] Figure 65. Shows that the phage, hybridoma and BM anti-human TIGIT antibodies, 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, as well as BM26, increase antigen-specific ΙΚΝγ signaling either alone (open bars) or in combination with an anti-PVRIG antibody, CHA.7.518.1.H4(S241P) (hatched bars). BM26 is the mlgGl isotype. For the isotype antibody control treatments, the open bar refers to the isotype antibody alone, and the hatched bar refers to isotype antibody in combination with

CHA.7.518.1.H4(S241P). Representative data (n=2) shows the amount of IFNy (mean +/- standard deviation) in the culture supernatant after a 24 hour co-culture of CMV-specific CD8+ T cells with Mel624 cells over-expressing human PVR and human PVRL2. The concentration of each antibody was 10 μg/ml. The Me 1624 human PVR/human PVRL2 cells used in the assay were pulsed with 0.0033 μg/ml or 0.001 μg/ml peptide.

[0095] Figure 66. Shows the percent increase of IFNy secretion with anti-human TIGIT antibodies, CHA.7.518.1.H4(S241P), and the combination of anti-human TIGIT antibodies and

CHA.7.518.1.H4(S241P), over the respective isotype control antibodies.

[0096] Figure 67 is the dendrogram for the epitope binning experiments of Example 7.

[0097] Figure 68 is the grouping of the antibodies from the epitope binning experiments of Example 7.

[0098] Figure 69. Shows the high affinity binding to human TIGIT overexpressing cells in a dose titration of the affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), benchmark antibodies (BM26, BM29), and the hIgG4 isotype control (anti-Synagis) on human TIGIT over-expressing Expi293 cells, as described in experiments of Example 3. All antibodies were titrated using a serial 2-fold dilution over 11 points starting at 10 μg/ml (133.33 nM [binding site]). AF647-labeled goat anti-human F(ab') (Jackson Immunoresearch) was added to the cells to detect binding of anti-TIGIT antibodies. The gMFI of the anti-TIGIT antibodies bound to the human TIGIT over-expressing Expi293 cells (black line), and the parental Expi293 cells (grey line) are shown. KD values +/- 95% CI, and curve fits are indicated below each graph.

[0099] Figure 70. Shows that anti-TIGIT antibodies are cross reactive to cyno TIGIT in a dose titration of the affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), benchmark antibodies (BM26, BM29), and the hIgG4 isotype control (anti-Synagis) on cyno TIGIT over-expressing Expi293 cells, as described in experiments of Example 3. All antibodies were titrated using a serial 2-fold dilution over 11 points starting at 10 μg/ml (133.33 nM [binding site]). AF647-labeled goat anti-human F(ab') (Jackson Immunoresearch) was added to the cells to detect binding of anti-TIGIT antibodies. The gMFI of the anti-TIGIT antibodies bound to the cyno TIGIT over-expressing Expi293 cells (black line), and the parental Expi293 cells (grey line) are shown. KD values +/- 95% CI, and curve fits are indicated below each graph.

[00100] Figure 71. Shows that affinity matured phage antibodies are cross reactive to mouse TIGIT in a dose titration of the affinity matured phage antibodies reformatted as mouse IgGl (mlgGl) (CPA.9.083, CPA.9.086), benchmark anti-mouse TIGIT antibodies (BM27 mlgGl, BM30 mlgGl), and the mlgGl isotype control (anti-Synagis) are shown, as described in experiments of Example 3. A) The gMFI of the anti-TIGIT antibodies bound to the mouse TIGIT over-expressing HEK cells (black line), and the parental HEK cells (grey line). B) The gMFI of the anti-TIGIT antibodies (black line) or Synagis mlgGl (grey line) bound to regulatory CD4+CD25+Foxp3+ T cells isolated from s.c. implanted Renca tumors in Balb/c mice. Anti-TIGIT antibodies were titrated using either a serial 2- or 3-fold dilution series starting at 15 μg/ml (200 nM [binding site]), or 10 μg/ml (132 nM [binding site]), respectively. AF647-labeled goat anti-mouse IgG-Fc (Southern Biotech) were added to the cells to detect binding of the anti-TIGIT antibodies on mouse TIGIT over-expressing cells. Anti-TIGIT antibodies were directly conjugated to AF647 for mouse Treg binding. KD values for each anti-TIGIT antibody are indicated.

fOOlOl] Figure 72. Shows a dose titration of the affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), and benchmark antibodies (BM26, BM29) on human effector memory CD95+CD28-CD8+CD3+ T cells from 3 healthy donor PBMCs (Donors 321, 322, and 334), as described in experiments of Example 3.

PBMCs were surface stained with antibodies against the following lineage markers CD3, CD4, CD8, CD 14, CD 16, CD28, CD56, and CD95 (BD Biosciences, BioLegend), as well as live/dead fixable aqua dye (Life Technologies). AF647-labeled anti-TIGIT antibodies and hIgG4 isotype control antibody (anti-Synagis) were then titrated using a serial 3-fold dilution over 12 points starting at 30 μg/ml (396 nM [binding site]). The gMFI of the anti-TIGIT antibodies bound to the effector memory T cells are shown. KD values for each antibody across the 3 different donors are reported in the table. The affinity mature phage antibodies (CPA.9.083 and CPA.9.086) had the highest binding affinity to the human effector memory T cells.

[00102] Figure 73. Shows a dose titration of the affinity matured phage antibodies (CPA.9.083, CPA.9.086, CPA.9.103), humanized hybridoma antibody (CHA.9.547.1), and benchmark antibody (BM26) on cyno effector memory CD95+CD28"CD8+CD3+ T cells from PBMCs isolated from 2 naive cyno monkeys (BioreclamationlVT), as described in experiments of Example 3. PBMCs were surface stained with antibodies against the following lineage markers CD3, CD4, CD8, CD 14, CD 16, CD28, CD56, and CD95 (BD Biosciences, BioLegend), as well as live/dead fixable aqua dye (Life

Technologies). AF647-labeled anti-TIGIT antibodies and hIgG4 isotype control antibody (anti-Synagis) were then titrated using a serial 3-fold dilution over 12 points starting at 30 μg/ml (396 nM [binding site]). The gMFI of the anti-TIGIT antibodies bound to the effector memory T cells are shown with the gMFI of the anti-Synagis hIgG4 isotype control antibody subtracted. KD values for each antibody across the 2 donors are reported in the table. The affinity mature phage antibodies (CPA.9.083 and CPA.9.086) had the highest binding affinity to the cyno effector memory T cells.

[00103] Figure 74. Shows the SPR kinetics of anti-TIGIT antibody binding to human, cyno and mouse TIGIT, as described in experiments of Example 5. The kinetic rate and equilibrium dissociation constants for the 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 benchmark antibodies (BM26, BM29) were determined by SPR on the ProteOn instrument.

[00104] Figure 75. shows that the anti-TIGIT antibodies block PVR/TIGIT interactions, as described in experiments of Example 4. Human TIGIT over-expressing Expi293 cells were preincubated with either the affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), benchmark antibodies (BM26, BM29), or the hIgG4 isotype control (anti-Synagis). All antibodies were titrated using a serial 2.5-fold dilution over 11 points starting at 10 μg/ml (133.33 nM [binding site]). Following antibody preincubation, human PVR-m2aFc was added to the cells at 158 nM [binding site] or EC90. AF647-labeled goat anti-mouse IgG-Fc (Southern Biotech) was then added to the cells to detect binding of anti-TIGIT antibodies. The percent inhibition of PVR-m2aFc binding to the human TIGIT over-expressing Expi293 cells is shown for each antibody. IC50 values for each anti-human TIGIT antibody are reported in the table (n=2 experiments).

[00105] Figure 76. Show the results of Example 6, that the affinity matured phage antibodies

(CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), and benchmark antibody (BM26) increase IL-2 signaling in a dose-dependent manner. Synagis hIgG4 is the isotype control antibody. Representative data (n>2) shows the RLU (mean +/- standard deviation) of the luciferase signal from a 6 hour co-culture of Jurkat IL-2-RE luciferase human TIGIT cells and CHO-K1 human PVR cells. A 19 point, 1.5-fold dilution series starting at 20 μg/ml was used for each antibody.

[00106] Figure 77. Shows that anti-TIGIT antibodies induce IFNy in CMV-specific CD8+ T cells. An in vitro co-culture assay with human CMV-specific CD8+ T cells was utilized to assess the effect of the affinity matured phage antibodies (CPA.9.083, CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), and benchmark antibodies (BM26, BM29) on antigen-specific cytokine secretion, as described in experiments of Example 6. The target cell line used in the assay was the HLA-A2+ pancreatic adenocarcinoma cells, Pane.05.04 that endogenously expresses human PVR and PVRL2. Panc.05.04 cells were pulsed with the CMV pp65 peptide at 0.03 μg/ml or 0.01 μg/ml at 37 °C for 1 hour. Cells were then washed and plated at 50,000 cells/well in 96-well round-bottom tissue culture treated plates. Anti-human TIGIT antibodies or the isotype control hIgG4 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 protocol above. 50,000 human CD8+ T cells were added to each well. Co-cultures were incubated at 37 °C with 5% C02 for 24 hours. The amount of human interferon gamma (IFNy) in the co-culture supernatant was measured by flow cytometry

using a cytometric bead assay (BD Biosciences). The percent increase of IFNy secretion for each antibody over the hIgG4 isotype is summarized in the table (n=2 experiments).

[00107] Figure 78. Shows anti-TIGIT antibodies augment IFNy when combined with a PVRIG antibody, CHA.7.518.1.H4(S241P). An in vitro co-culture assay with human CMV-specific CD8+ T cells was utilized to assess the effect of the affinity matured phage antibodies (CPA.9.083,

CPA.9.086), humanized hybridoma antibodies (CHA.9.547.7, CHA.9.547.13), and benchmark antibodies (BM26, BM29) on antigen-specific cytokine secretion in combination with an anti-PVRIG antibody, CHA.7.518.1. The target cell line used in the assay was the HLA-A2+ pancreatic adenocarcinoma cells, Pane.05.04 that endogenously expresses human PVR and PVRL2. Pane.05.04 cells were pulsed with the CMV pp65 peptide at 0.03 μg/ml or 0.01 μg/ml at 37 °C for 1 hour. Cells were then washed and plated at 50,000 cells/well in 96-well round-bottom tissue culture treated plates. Anti-human TIGIT antibodies or the isotype control hIgG4 antibody (anti-Synagis) were added at a concentration of 0.1 μg/ml in combination with CHA.7.518.1 (hatched bars) or a control hIgG4 isotype antibody at 10 μg/ml (solid bars). Human CMV-specific CD8+ T cells from a single donor were expanded according to the protocol above. 50,000 human CD8+ T cells were added to each well. Co-cultures were incubated at 37 °C with 5% C02 for 24 hours. The amount of human IFNy in the co-culture supernatant was measured by flow cytometry using a cytometric bead assay (BD

Biosciences). The percent increase of IFNy secretion for each antibody over the hIgG4 isotype is summarized in the table (n=2 experiments).

[00108] Figure 79. Shows the correlation analysis of PVRIG and TIGIT expression on CD4+ and CD8+ T cells from dissociated tumors. For each tumor sample, a mean flourescence intensity ratio (MFIr) was calculated, and a Spearman's correlation analysis was performed, and an r2 and p value reported.

[00109] Figure 80. Shows the results of tumor growth inhibition and survival in TIGIT KO mice treated with an anti-mouse PVRIG antibody. Groups of 7-10 TIGIT KO and C57BL/6 WT mice were s.c. injected with 1 χ 105 B16/Db-hmgpl00 cells. Mice were treated twice per week for 3 weeks, starting at the inoculation day (day 0) with the designated antibody. A) Mean tumor volumes +/-standard error of the mean (SEM) are shown in the upper graph, with *** indicating a p-value <0.001 for TIGIT KO treated with anti-mouse PVRIG antibody (Clone 407) compared to C57BL/6 WT treated with the mlgGl isotype control antibody. Tumor volumes for individual mice within each antibody treatment group are shown as spider plots in lower graphs. B) Table summarizing the TGI as measured at indicated days compared to control C57BL/6 WT mice treated with the mlgGl isotype control. C) Survival of mice after s.c. injection of B16/Db-hmgpl00 cells.

[00110] Figure 81A to C depicts combination treatments with the indicated antibodies as compared to control in Mel-624, Colo205, and Pane.05.04 cells. gplOO or CMVpp65 specific T cells were co-cultured with Mel-624, Colo205, and Pane.05.04 cells, gplOO or CMVpp65 peptide, and the indicated antibodies at 10 mg/ml. IFN-γ concentration in the conditioned media was determined at 24 hrs. Average + Std Dev of triplicates is shown. % change in IFN-γ for each condition relative to hIgG4 is shown.

[00111] Figure 82 A to C depict expression of PD-1/TIGIT/PVRIG on CD8 T cells and expression of PD-L1, PVR, PVRL2 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. Expression of PVRIG, TIGIT, and PD-1 on CMVpp65 reactive T cells is shown. B) Expression of PD-L1, PVR, and PVRL2 on Colo205 and Panc.05.04 cells is shown. C) CMVpp65 specific T cells were co-cultured with Colo205 and Panc.05.04 cells, CMVpp65 peptide, and the indicated antibodies at 10 mg/ml. IFN-γ concentration in the conditioned media was determined at 24 hrs. Average + Std Dev of triplicates is shown. % change in IFN-γ for each condition relative to hIgG4 is shown.

[00112] Figure 83. PVRIG is expressed highest on cytotoxic lymphocyte subsets from human cancer. A) Expression of PVRIG on leukocyte cell subsets from 5-8 healthy donor PBMCs is shown. PVRIG expression is defined as the ratio of PVRIG MFI relative to isotype control MFI. B) Expression of PVRIG, TIGIT, CD96, and PD-1 on peripheral blood Tregs as compared to CD8 T cell subsets from 5 healthy donor PBMCs is shown. C) CMV pp65 specific T cells from 3 healthy donors were expanded in vitro with pp65 (495 - 503) peptide, IL-2 and IL-7 for up to 7 days. Expression of TIGIT (blue) and PVRIG (black) on HLA-A2/pp65 (495 - 503) tetramer positive cells is shown. D) Human T cells were cultured with allogeneic DCs and expression of TIGIT and PVRIG shown on CD4+ T cells on day 0, 1, 2, and 7 post activation. E) Representative FACS plots showing expression of PVRIG (blue) compared to isotype control (red) on TILS (CD4 T cells, CD8 T cells, and NK cells) from a representative lung and kidney cancer. F) Co-expression of PVRIG, TIGIT, and PD-1 on CD4 and CD8 TILS from a lung cancer sample is shown. G) Expression of PVRIG on CD8+ and CD4+ TILS from dissociated human tumors of various cancer types is shown. Each dot represents a distinct tumor from an individual patient. H) Relative expression on CD8 TILs vs Treg TILS for PVRIG, TIGIT, and PD-1 from endometrial, kidney, and lung tumors was assessed. For each tumor, the fold expression on CD8 TILS was normalized to fold expression on Treg TILS and plotted. For A, B, C, G, and H, mean + SEM is shown by the error bars.

[00113] Figure 84. PVRL2 expression is enhanced in the tumor microenvironment. A) PVRL2 expression was assessed by IHC on lung, ovarian/endometrial, breast, colon, kidney, and melanoma tumors. Bars depict mean + SEM. For each tumor, 2 cores were assessed by a

pathologist and scored based on prevalence and intensity of membranous staining on tumor cells as described in the supplemental methods. For each tumor, the average score of 2 cores is shown. B) A representative melanoma tumor showing PVRL2 expression on tumor cells (arrow) and in the immune cells (*) in the stroma is shown. C) PVRL2 expression on a log2 scale from dissociated tumors determined by FACS on CD45", CD14+ TAMs, and Lin"CD14" CD33W mDC cell subsets is shown. Mean + SEM is shown for each cancer type. Dotted line represents no staining was observed. For each cell type, at least 100 events were required in order to be analyzed. D) Representative FACS plots for PVRL2 expression (blue) as compared to IgG (red) are shown for a lung cancer. E) For tumor samples where we were able to assess both PVRIG and PVRL2 expression, PVRIG expression on CD8+ T cells is plotted versus PVRL2 expression on CD14+ TAMS and CD45" cells for each tumor. Each dot represents an individual tumor sample. Red line represents a 2 fold expression of PVRIG or PVRL2 compared to IgG. The Table in Figure 84F shows the prevalence of PVRL2 in various tumor samples.

|00114] Figure 85. Distinct regulation of PVRL2 and PD-Ll on tumor cells. A) Expression of PD-Ll and PVRL2 was assessed by IHC on serial sections. Tumors samples from Figure 84 A were grouped based on tissue type and expression of PVRL2 on PD-Ll negative and PD-Ll positive is shown. PD-Ll negative tumors were defined as no membranous staining on tumor or immune cells from either duplicate cores for a given tumor. PD-Ll positive staining was defined as membranous staining on at least 1 core of a tumor. Bars depict mean + SEM for each group. B, C) Representative expression of a PVRL2+PD-L1" endometrial (B) tumor and a PVRL2+PD-L1" lung (C) tumor. D) Immature BM-DCs were cultured with the indicated stimuli and PVR, PVRL2, and PD-Ll expression assessed by FACS on day 2 of culture. For each condition, expression was normalized to media only control condition. E) Expression of PVR, PVRL2, and PD-Ll on HT-29 cells treated with IFN-D or media alone is shown. PD-Ll or PVRL2 is shown in blue and IgG isotype control staining is shown in red.

[00115] Figure 86. CHA.7.518.1.H4(S241P) is a high affinity antibody that enhances T cell activation. A) Binding of CHA.7.518.1.H4(S241P) or IgG isotype control to HEK293 PVRIG or HEK293 parental cells by FACS is shown. FACS KD values are shown for the binding of

CHA.7.518.1.H4(S241P) to HEK293 hPVRIG, HEK293 cPVRIG, and Jurkat cells. B)

CHA.7.518.1.H4(S241P) disrupts the binding of PVRL2 Fc to HEK293 cells ectopically expressing PVRIG. Mean + Std Dev of triplicate values is shown. C) CHA.7.518.1.H4(S241P) blocks the binding of PVRIG Fc to HEK293 cells that endogenously express PVRL2. D) Human CD4 T cells were co-cultured with aAPC CHO cells expressing a cell surface bound anti-CD3 antibody and hPVRL2 in the presence of 10 μg/ml anti-PVRIG antibody and human IgG isotype control antibodies. The effect of anti-PVRIG Ab on proliferation of CD4 T cells isolated from 11 different donors is shown. Bars depicted mean + SEM. E) gplOO specific T cell lines (TIL-209, TIL-463) were co-cultured with CHO cells engineered to express HLA-A2 and PVRL2 along with 10 μg/ml anti-PVRIG or IgG isotype control antibody. IFN-γ and TNF-a production was tested at 24 hours post co-culture. Mean + Std Dev of triplicate values is shown. Percent change in IFN-γ and TNF-a for each condition relative to isotype control is depicted by the 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 is shown. For the T cells, expression of PVRIG, TIGIT, and PD-1 (red) relative to IgG (blue) on TIL-209 and TIL-463 gplOO specific T cells, and on CMVpp65 specific T cells is shown. To expand 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 is shown on HLA-A2/pp65 tetramer positive cells. G) gplOO specific T cells (TIL-209, TIL-463) expanded from TILS derived from melanoma tumors were co-cultured with MEL624 cells in the presence of 10 Dg/ml of the indicated antibodies. IFN-γ concentration in the conditioned media was determined at 24hrs. H, I) Expanded CMVpp65 specific T cells were co-cultured with Colo205 and Panc.05.04 cells, CMVpp65 peptide, and the indicated antibodies at 10 □ g/ml. IFN-γ concentration in the conditioned media was determined at 24 hrs. For E, G, H, I, average + Std Dev of triplicates is shown. Percent change in IFN-γ for each condition relative to isotype control is depicted by the number above each bar.

[00116] Figure 87. PVRIG deficient mice have increased T cell function. A) RNA expression of PVRIG as measured by qRT-PCR from purified mouse immune cell subsets was assessed. Relative expression to housekeeping was determined by DCt method. B) pmel CD8+ TCR transgenic T cells were activated with gplOO (25-33) and PVRIG and TIGIT RNA transcript levels assessed by qRT-PCR at the indicated time points. Graph shows mean + SEM of results from 5 different experiments. C) Spleens were harvested from PVRIG"7" and WT littermates and analyzed by flow cytometry for expression of PVRIG 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. Following the activation, PVRIG expression on CD4+ and CD8+ T cells ("activated" cells) was analyzed by flow cytometry. Each dot represents cells derived from an individual mouse. D) WT and PVRIG" " derived splenocytes were labeled with Cell Proliferation Dye eFluor450 and were cultured in the presence of Control-Fc (mouse IgG2a) or with mouse PVRL2 Fc. After 4 d of culture, cell division was analyzed by flow cytometry. Representative FACS plots from an experiment (left) and the summary of percentage inhibition by PVRL2 Fc (defined as % proliferation Control-Fc subtracted from % proliferation PVRL2 Fc) 3 independent experiments (right) are presented. * indicate p-value < 0.05, paired student's t-test for the change in proliferation in the presence of PVRL2-FC relative to proliferation in the presence of protein control in WT versus PVRIG" " T cells. E) pmel CD8+ T cells derived from pmel PVRIG"7" or pmel PVRIG WT mice were activated for 11 days with their cognate peptide and IL2. Activated pmel CD8+ cells were then co-cultured with B16-Db/gpl00 cells for 18 hours and following the co-culture were evaluated for CD 107 expression and for cytokine production. Four independent experiments are presented as indicated by each paired dot. * indicate p-value<0.05, Student's t-test comparing PVRIG"7" versus WT.

[00117] Figure 88. PVRIG deficiency results in reduced tumor growth and increased CD8+ effector T cell mechanism. A) C57BL/6 WT or PVRIG" " mice were subcutaneously injected with 5 * 105 MC38 cells. Tumor volumes were measured x2 weekly. n= 10 mice per group, Ave + SEM is shown, * Indicate p-value<0.05 by Student's unpaired t-test for WT mice versus PVRIG"7" mice (ANOVA). B) Individual tumor growth curves are shown. n=10 mice per group, one representative experiment is shown (n=2). C) C57BL/6 WT or PVRIG"7" mice were subcutaneously injected with 5* 105 MC38 cells. At day 14 post- inoculation, mice were treated with anti-PD-Ll, x2 weekly for 2 weeks. Tumor volumes were measured x2 weekly. n=10 mice per group, Ave + SEM is shown, p-value= 0.052 by Student's unpaired t-test for WT mice versus PVRIG"7" mice, both treated with anti-PD-Ll. D) Individual tumor growth curves are shown. One representative experiment is shown (n=2). E-H) In separate duplicate experiments, tumors were harvested on day 18 after mice had received 2 doses of anti-PD-Ll or the relevant isotype control. Dissociated tumors were enriched for CD45+ cells prior to stimulation for 4 hours with PMA and Ionomycin in the presence of Brefeldin A. Graphs illustrate the total numbers per mg tumor tissue of CD45+ immune cells, CD8+ T cells and Interferon-y-producing CD8+ T cells from isotype-treated wild-type and PVRIG"7" mice (E) and from anti-PD-Ll -treated wild-type and PVRIG"7" mice (F). G-H) Frequency of CD8+ IFN-D+ TNF-D+ effector cells in tumor-draining lymph nodes from isotype- and anti-PD-Ll -treated PVRIG"7" mice, relative to their corresponding wild-type cohort is shown. For E-H, Ave + SEM is shown and p values from a Student's unpaired t-test is shown.

[00118] Figure 89. Antagonistic anti-PVRIG antibodies synergistically inhibit tumor grown in combination of PD-1 inhibitors or TIGIT genetic deficiency. A) Binding of mPVRL2 Fc fusion protein to mPVRIG HEK293 engineered cells that were pre-incubated with serial dilutions of anti-mPVRIG mAb or IgG isotype control Ab is shown. B) BALB/c mice were subcutaneously injected with 5x 105 CT26 cells. On day 14 post inoculation, mice were sacrificed and spleen, draining lymph nodes and tumors were harvested. Cells were analyzed by flow cytometry for expression of PVRIG on CD3+CD4+ T cells, CD3+CD8+ T cells, CD3"CD49b+ NK cells, CD l lb+ Gr- Myeloid-Derived-Suppressor Cells (MDSC) and CDl lb+F4/80+ macrophages. C,D) BALB/c mice were subcutaneously injected with 5 χ 105 CT26 cells. At day 7 post inoculation mice were treated with anti-PD-Ll and/or anti-PVRIG Ab, 2x weekly for 3 weeks (arrows indicate Ab treatment). C) Tumor volumes are shown. *** indicate p-value < 0.001 (ANOVA) for aPD-Ll+Rat IgG2b compared to DPD-Ll+aPVRIG treated groups. Arrows indicate when antibodies were dosed. D. Survival analysis of complete responder's mice. * indicate p value < 0.05 (Log-rank test) for DPD-Ll + Rat IgG2b compared to DPD-Ll + DPVRIG treated groups. One representative study of 3 studies are shown. E. C57BL/6 or TIGIT"7" mice were subcutaneously injected with 1 χ 105 B16/Db-hmgpl00 cells. Mice were treated 2x weekly for 3 weeks with the designated mAb starting on the day of inoculation (day 0). E. Tumor volumes were measured 2x weekly and average + SEM is shown shown. Tumor growth inhibition as measured at indicated days compared to control WT+mlgGl isotype control. *** indicate p-value<0.001 for TIGIT"7" + aPVRIG compared to WT + mlgGl isotype control. Arrows indicate when antibodies were dosed. F. Individual tumor growth curves for each mouse is shown. One representative experiment out of 2 performed is shown.

[00119] Figure 90. PVRIG is expressed on T and NK cells of TILS in human cancer. A) Expression of PVRIG, TIGIT, CD96, and PD-1 on CD4 T cell subsets from healthy donor PBMCs is shown. Mean + SEM is shown. B) Human T cells were co-cultured with allogeneic PBMCs and expression of PVRIG protein on CD4 and CD8 T cells shown (top). C) Tumors were dissociated and single cells were activated with anti-CD3 and anti-CD28. Expression of PVRIG (blue) relative to IgG isotype control (red) was assessed on day 0 (directly ex vivo) and day 5 post activation. D) Expression of PVRIG on NK cells from dissociated human tumors is shown. Each dot represents a distinct tumor from an individual patient. Mean + 95% confidence internal is shown. D) Dissociated tumor cells were activated with 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) Expression of PVRIG was assessed on CD4 and CD8 T cells from dissociated tumors and from dissociated donor-matched normal adjacent tissue. Each line represents matched tissues obtained from an individual patient. A paired student's t-test was performed. F) A correlation analysis of the magnitude of PVRIG, TIGIT, and PD-lfold expression relative to IgG isotype control on CD4 and CD8 T cells from tumors is shown. Each dot represents an individual tumor sample. A Spearman's correlation coefficient and p value are shown.

[00120] Figure 91. Expression of PVRL2 is enhanced in colon, skin, and breast cancers. A) Photomicrographs showing the binding of Sigma anti human PVRL2 antibody to FFPE sections of positive cells, CHO-S human PVRL2 (right) compare to negative cells, CHO-S (left), following antigen retrieval 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) Example expression of PVRL2 in lung normal and cancer tissues. C) Normal tissue showing no staining. D) Lung Adenocarcinoma showing partial positive staining. E) Lung adenocarcinoma showing positive staining. F) Lung adenocarcinoma showing strong positive staining.

[00121] Figure 92. PVRL2 is upregulated on TAMs and CD45" cells in the tumor as compared to normal adjacent tissue. Expression of PVRL2 on CD45" cells and TAMs from donor matched tumor and normal adjacent tissue is shown. A paired student's t-test p value is shown.

[00122] Figure 93. PVRIG and PVRL2 are co-expressed in the same tumor sample. PVRIG expression on CD4 T cells (A) and NK cells (B) is plotted against PVRL2 expression on TAMS for an individual tumor.

[00123] Figure 94. Activity of CHA.7.518.1.H4(S241P) on human T cells. A) Expression of

PVRIG on CD4 T cells activated with CHO cells expressing cell surface bound anti-CD3 and PVRL2. B) Expression of HLA-A2, B-2m, and PVRL2 are shown on CHO-S parental and engineered CHO-S cell lines. Fold expression relative to isotype is depicted by the 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 varying concentrations of anti-PVRIG Ab or relevant IgG control. % Proliferation is shown. Each dot represents an average of triplicate values. D) CHO cells ectopically expression HLA-A2/B2m and PVRL2 were co-cultured with 2 gplOO specific T cell lines (TIL F4, TIL 209) in the presence of 1 ug/ml gplOO and varying concentrations of anti-PVRIG antibody or relevant IgG control. TNF-a concentrations on day 3 of co-culture is down. Each value represents an average of triplicates.

[00124] Figure 95. Characterization of mPVRIG binding interactions and a surrogate anti-mPVRIG antibody. A, B) Binding of mPVRIG to mPVRL2 was assessed by surface plasmon resonance. C) Soluble receptor Fc or control proteins were incubated in a dose response with immobilized mPVRL2 HIS in an ELISA format. Bound receptor Fc is shown. D) Soluble PVRL2 HIS protein was incubated in a dose response with PVRIG Fc or DNAM Fc coated plates. E) Binding of mPVRIG Fc or control Fc fusion protein to B16-F10 cell line transfected with mPVRL2 siRNA, mPVRsRNA, or scrambled siRNA transfection is shown. F) Affinity characterization of rat anti-mouse PVRIG mAb was performed by examining the binding of anti-mPVRIG to HEK293 cells overexpressing mPVRIG. G) Affinity characterization of rat anti-mouse PVRIG mAb was performed by examining the of anti-mPVRIG to D10.G4.1 cell line endogenously expressing mPVRIG vs isotype control rat IgG is shown. H) Binding of anti-mPVRIG to D 10.G4.1 cells transfected with mouse PVRIG-siRNA (green histogram) vs scr siRNA (orange histogram). I) Binding of mPVRIG Fc pre-incubated with anti-mPVRIG Ab to B16-F10 cells, which endogenously express PVRL2

[00125] Figure 96. Generation of transgenic PVRIG and TIGIT knockout mice. The PVRIG conditional knockout and Tigit knockout mouse lines were generated by Ozgene Pty Ltd (Bentley WA, Australia). A) The targeting construct in which PVRIG exons 1 to 4 were floxed was electroporated into a C57BL/6 ES cell line, Bruce4 (Roentgen et al., Int Immunol 5: 957-964, 1993). B) The targeting construct in which the coding region of Tigit exon 1 (including the ATG) and exons 2 and 3 were replaced with an FRT-flanked neo cassette was electroporated into a C57BL/6 ES cell line, Bruce4. Homologous recombinant ES cell clones were identified by Southern hybridization and injected into goGermline blastocysts (Roentgen et al., genesis 54: 326-333, 2016). Male chimeric mice were obtained and crossed to C57BL/6J females to establish heterozygous germline offspring on C57BL/6 background. The germline mice were crossed to a ubiquitous FLP C57BL/6 mouse line to remove the FRT flanked selectable marker cassette and generate the conditional or knockout alleles (for PVRIG and Tigit, respectively). For PVRIG knockout, mice were further crossed to a ubiquitous Cre C57BL/6 mouse line to remove the loxP flanked exons and generate the knockout allele.

[00126] Figure 97. PVRIG knockout mice are immune-phenotypically similar to wild-type mice. Mice (n= 5 per wild-type and PVRIG knockout cohorts) were euthanized prior to venous blood being collected in anti-coagulant-coated tubes and harvesting of organs. Single cells were recovered from freshly harvested bone marrow, thymus, spleen, cutaneous and mesenteric lymph nodes. Cells were stained with fluorochrome-conjugated surface marker antibodies and acquired on a BD LSR Fortessa flow cytometer. Panels illustrate 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) across lymphoid tissue types. (H-I) Whole venous blood was run on a Hemavet 950 veterinary hematology system to compare differential counts and frequencies of blood cell subsets from wild-type and PVRIG deficient mice.

[00127] Figure 98. Increased T cell effector function in PVRIG"7" mice treated with anti-PDLl compared to WT with anti-PD-Ll. MC38 tumors were inoculated into WT or PVRIG"7" mice and were subsequently treated with anti-PD-Ll or rat IgG2b isotype control. On day 18, CD45+ tumor infiltrating lymphocytes were purified from tumors, RNA extracted, and transcript profiling performed. Several T cell related genes are shown, with each dot representing an individual mouse. Student's t test p values are shown.

[00128] Figure 99. Anti-TIGIT and anti-PVRIG antibodies induce tumor cell killing. An in vitro co-culture assay with human CMV-specific CD8+ T cells expanded was utilized to assess the effect of the benchmark anti-TIGIT antibody and CHA.7.518. l.H4(S241P) on antigen-specific tumor cell killing. HLA-A2+ target cell lines used in the assay were the Mel624 (A) and Panc05.04 (B). Synagis hIgG4 is the isotype control antibody. Luciferase activity in the target cells was measured with the Bio-Glo luciferase substrate. Representative data (n>2) shows the percent specific killing (mean +/- standard deviation) of Mel624 or Panc05.04 cells after a 16 hour co-culture with human CMV-specific CD8+ T cells from three different donors.

[00129] Figure 100. Dose-dependent tumor cell killing of anti-TIGIT antibodies with

CHA.7.518.1.H4(S241P). An in vitro co-culture assay with human CMV-specific CD8+ T cells was utilized to assess the effect of two different anti-TIGIT antibodies, BM26 and CPA.9.086 when combined with CHA.7.518.1.H4(S241P) on antigen-specific Mel624 cell killing. Luciferase activity in the target cells was measured with the Bio-Glo luciferase substrate. Representative data (n>2) shows the percent specific killing (mean +/- standard deviation) of Mel624 cells after a 16 hour co-culture with human CMV-specific CD8+ T cells from one donor.

[00130] Figure 101. CPA.9.086 CDR sequences, IMGT and Kabat numbering.

[00131] Figure 102. Anti-TIGIT hIgG4 + CHA.7.518.1.H4(S241P) combination induces tumor cell killing. Co-culture of CMV-reactive CD8+ T cells with Mel624 PVR, PVRL2 & luciferase OE Single dose of 10 μ^ιηΐ aTIGIT Ab and 10 μ^ιηΐ CHA.7.518.1.H4(S241P) with CMV-reactive donor 4, while dose titration starting at 0.5 μ^ιηΐ aTIGIT Ab and 10 μ^ιηΐ CHA.7.518.1.H4(S241P) with CMV-reactive donor 156.

V. DETAILED DESCRIPTION OF THE INVENTION

A. Overview

[00132] The present invention provides a number of useful antibodies, for use alone or in combination, for treatment of cancer. Cancer can be considered as an inability of the patient to recognize and eliminate cancerous cells. In many instances, these transformed (e.g. cancerous) cells counteract immunosurveillance. There are natural control mechanisms that limit T-cell activation in the body to prevent unrestrained T-cell activity, which can be exploited by cancerous cells to evade or suppress the immune response. Restoring the capacity of immune effector cells-especially T cells-to recognize and eliminate cancer is the goal of immunotherapy. The field of immuno-oncology, sometimes referred to as "immunotherapy" is rapidly evolving, with several recent approvals of T cell checkpoint inhibitory antibodies such as Yervoy, Keytruda and Opdivo. These antibodies are generally referred to as "checkpoint inhibitors" because they block normally negative regulators of T cell immunity. It is generally understood that a variety of immunomodulatory signals, both costimulatory and coinhibitory, can be used to orchestrate an optimal antigen-specific immune response. Generally, these antibodies bind to checkpoint inhibitor proteins such as CTLA-4 or PD-1, which under normal circumstances prevent or suppress activation of cytotoxic T cells (CTLs). By inhibiting the checkpoint protein, for example through the use of antibodies that bind these proteins, an increased 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, leading to immune stimulation, resulting in treatment of conditions such as cancer and infectious disease.

[00133] The present 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 to other known immune checkpoints. The identification and methods used to show that PVRIG is a checkpoint receptor are discussed in WO2016/134333, expressly incorporated herein by reference. Antibodies to human PVRIG that block the interaction and/or binding of PVLR2 are provided herein. When PVRIG is bound by its ligand (PVRL2), an inhibitory signal is elicited which acts to attenuate the immune response of NK and T-cells against a target cell (i.e. analogous to PD-1/PDL1). Blocking the binding of PVRL2 to PVRIG shuts-off this inhibitory signal of PVRIG and as a result modulates the immune response of NK and T-cells. Utilizing an antibody against PVRIG 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 generated which bind PVRIG and block the binding of its ligand, PVRL2. Anti-PVRIG antibodies in combination with other checkpoint inhibitor antibodies such as PD-1 are provided.

[00134] Similarly, TIGIT has been shown to also have attributes of a checkpoint receptor, and the present invention provides anti-TIGIT antibodies that block the interaction and/or binding of TIGIT to PVR are provided. When TIGIT is bound by its ligand (PVR), an inhibitory signal is elicited which acts to attenuate the immune response of NK and T-cells against a target cell (i.e. analogous to PD-1/PDL1). Blocking the binding of PVR to TIGIT shuts-off this inhibitory signal of TIGIT and as a result modulates the immune response of NK and T-cells. Utilizing an antibody against TIGIT that blocks binding to PVR is a therapeutic approach that enhances the killing of cancer cells by NK and T-cells. Blocking antibodies have been generated which bind TIGIT and block the binding of its ligand, PVR. Anti-TIGIT antibodies in combination with other checkpoint inhibitor antibodies such as PD-1 are provided.

[00135] Additionally, the invention provides combinations of anti-PVRIG and anti-TIGIT antibodies for use in the treatment of cancer.

B. Definitions

[00136] In order that the application may be more completely understood, several definitions are set forth below. Such definitions are meant to encompass grammatical equivalents.

[00137] By "ablation" herein is meant a decrease or removal of activity. In some

embodiments, it is useful to remove activity from the constant domains of the antibodies. Thus, for example, "ablating FcyR binding" means the Fc region amino acid variant has less than 50% starting binding as compared to an Fc region not containing the specific variant, with less than 70-80-90-95-98% loss of activity being preferred, and in general, with the activity being below the level of detectable binding in a Biacore assay. As shown in Figure 50, one ablation variant in the IgGl constant region is the N297A variant, which removes the native glycosylation site and significantly reduces the FcyRIIIa binding and thus reduces the antibody dependent cell-mediated cytotoxicity (ADCC).

[00138] By "antigen binding domain" or "ABD" herein is meant a set of six Complementary

Determining Regions (CDRs) that, when present as part of a polypeptide sequence, specifically binds a target antigen as discussed herein. Thus, a "TIGIT antigen binding domain" binds TIGIT antigen (the sequence of which is shown in Figure 51) as outlined herein. Similarly, a "PVRIG antibody binding domain" binds PVRIG antigen (the sequence of which is shown in Figure 1) as outlined herein. As is known in the art, these CDRs are generally present as a first set of variable heavy CDRs (vhCDRs or VHCDRs) and a second set of variable light CDRs (vlCDRs or VLCDRs), each comprising three CDRs: vhCDRl, vhCDR2, vhCDR3 for the heavy chain and vlCDRl, vlCDR2 and vlCDR3 for the light. The CDRs are present in the variable heavy and variable light domains, respectively, and together form an Fv region. Thus, in some cases, the six CDRs of the antigen binding domain are contributed by a variable heavy and variable light chain. In a "Fab" format, the set of 6 CDRs are contributed by two different polypeptide sequences, the variable heavy domain (vh or VH; containing the vhCDRl, vhCDR2 and vhCDR3) and the variable light domain (vl or VL;

containing the vlCDRl, vlCDR2 and vlCDR3), with the C-terminus of the vh domain being attached to the N-terminus of the CHI domain of the heavy chain and the C-terminus of the vl domain being attached to the N-terminus of the constant light domain (and thus forming the light chain).

[00139] By "modification" herein is meant an amino acid substitution, insertion, and/or deletion in a polypeptide sequence or an alteration to a moiety chemically linked to a protein. For example, a modification may be an altered carbohydrate or PEG structure attached to a protein. By "amino acid modification" herein is meant an amino acid substitution, insertion, and/or deletion in a polypeptide sequence. For clarity, unless otherwise noted, the amino acid modification is always to an amino acid coded for by DNA, e.g. the 20 amino acids that have codons in DNA and RNA.

[00140] By "amino acid substitution" or "substitution" herein is meant the replacement of an amino acid at a particular position in a parent polypeptide sequence with a different amino acid. In particular, in some embodiments, the substitution is to an amino acid that is not naturally occurring at the particular position, either not naturally occurring within the organism or in any organism. For example, the substitution N297A refers to a variant polypeptide, in this case an Fc variant, in which the asparagine at position 297 is replaced with alanine. For clarity, a protein which has been engineered to change the nucleic acid coding sequence but not change the starting amino acid (for

example exchanging CGG (encoding arginine) to CGA (still encoding arginine) to increase host organism expression levels) is not 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 that it started with, it is not an amino acid substitution.

[00141] By "amino acid insertion" or "insertion" as used herein is meant the addition of an amino acid sequence at a particular position in a parent polypeptide sequence. For example, -233E or 233E designates an insertion of glutamic acid after position 233 and before position 234.

Additionally, -233ADE or A233ADE designates an insertion of AlaAspGlu after position 233 and before position 234.

[00142] By "amino acid deletion" or "deletion" as used herein is meant the removal of an amino acid sequence at a particular position in a parent polypeptide sequence. For example, E233- or E233#, E233() or E233del designates a deletion of glutamic acid at position 233. Additionally, EDA233- or EDA233# designates a deletion of the sequence GluAspAla that begins at position 233.

CLAIMS
1. A composition comprising an antigen binding domain that binds to human TIGIT (SEQ ID

NO:97) comprising:

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

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

2. A composition according to claim 1 wherein said composition is an antibody comprising: a) a heavy chain comprising VH-CHl-hinge-CH2-CH3, wherein said VH comprises SEQ ID NO: 160; and

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

3. A composition according to claim 2 wherein the sequence said CHl-hinge-CH2-CH3 is selected from human IgGl, IgG2 and IgG4, and variants thereof.

4. A composition according to claim 2 or 3 wherein said heavy chain has SEQ ID NO: 164 and said light chain has SEQ ID NO: 169.

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

6. A composition according to claim 5 wherein said second antibody binds human PD-1.

7. A composition according to claim 5 wherein said second antibody binds human PVRIG

(SEQ ID NO:2).

8. A composition according to claim 7 wherein said second antibody comprises an antigen binding domain comprising a variable heavy domain comprising SEQ ID NO: 5 and a variable light domain comprising SEQ ID NO: 10.

9. A composition according to claim 7 wherein 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. A nucleic acid composition comprising:

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

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

11. A nucleic acid composition according to claim 10 wherein said first nucleic acid encodes a heavy chain comprising VH-CHl-hinge-CH2-CH3, wherein said VH comprises SEQ ID NO: 160; and said second nucleic acid encodes a light chain comprising VL- VC, wherein said VL comprising SEQ ID NO: 165 and VC is the lambda domain.

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

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

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

15. A method of making an anti-TIGIT antibody comprising:

a) culturing said host cell of claim 14 under conditions wherein said antibody is expressed; and

b) recovering said antibody.

16. A method of treating cancer by activating T cells comprising administering an

composition according to any of claims 1 to 9.