Targeting novel inhibitory receptors in cancer immunotherapy

Abstract

T cells play a critical role in promoting tumor regression in both experimental models and humans. Yet, T cells that are chronically exposed to tumor antigen during cancer progression can become dysfunctional/exhausted and fail to induce tumor destruction. Such tumor-induced T cell dysfunction may occur via multiple mechanisms. In particular, immune checkpoint inhibitory receptors that are upregulated by tumor-infiltrating lymphocytes in many cancers limit T cell survival and function. Overcoming this inhibitory receptor-mediated T cell dysfunction has been a central focus of recent developments in cancer immunotherapy. Immunotherapies targeting inhibitory receptor pathways such as programmed cell death 1 (PD-1)/programmed death ligand 1 and cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), alone or in combination, confer significant clinical benefits in multiple tumor types. However, many patients with cancer do not respond to immune checkpoint blockade, and dual PD-1/CTLA-4 blockade may cause serious adverse events, which limits its indications. Targeting novel non-redundant inhibitory receptor pathways contributing to tumor-induced T cell dysfunction in the tumor microenvironment may prove efficacious and non-toxic. This review presents preclinical and clinical findings supporting the roles of two key pathways—T-cell immunoglobulin and mucin-domain containing-3 (TIM-3) and T cell immunoreceptor with Ig and ITIM domain (TIGIT)/CD226/CD96/CD112R—in cancer immunotherapy.

1. Introduction

Inhibitory receptors (IRs) on T cells regulate innate and adaptive immunity in chronic viral infections and cancer [1–4]. T cells exposed to chronic antigen stimulation can become dysfunctional/exhausted and upregulate many IRs, including programmed cell death receptor 1 (PD-1), T cell immunoglobulin mucin-3 (TIM-3), and T cell immunoreceptor with immunoglobulin (Ig) and ITIM domain (TIGIT) [2, 5–7]. At the same time, IR ligands are expressed by tumor cells and antigen-presenting cells (APCs) in the tumor microenvironment (TME). Targeting IRs with monoclonal antibodies (mAbs) has proven beneficial in mouse tumor models and humans, and immune checkpoint blockade (ICB) with anti-cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), anti-PD-1, or both mAbs are standard treatments for many solid tumors [8–11]. However, most patients with cancer do not respond to current ICB, and dual PD-1/CTLA-4 often causes serious grade adverse events that often result in treatment discontinuation [12–14]. These results prompt a need to identify additional IR pathways for therapeutic targeting. Multiple lines of evidence indicate that the TIM-3 and TIGIT/CD226/CD96/CD112R pathways limit adaptive and innate immunity against tumors [15–20]. Further, preclinical evidence in experimental mouse tumor models suggests that dual PD-1/TIM-3 and PD-1/TIGIT blockade can confer clinical benefit. Here, we review the preclinical and clinical findings supporting the promising role of TIM-3 and TIGIT/CD96/CD112R pathways in cancer immunology.

2. TIM-3

TIM-3 (CD366, HAVCR2) was originally identified as a transmembrane protein expressed by interferon gamma (IFN-γ)-producing CD4⁺ T helper type 1 (Th1) and CD8⁺ cytotoxic T cells [21]. TIM-3 is a member of the TIM gene family, which includes eight members (Tim 18) on mouse chromosome band 11B1.1, and three members (TIM-1/Havcr1, TIM-3/Havcr2, and TIM-4) on human chromosome 5q33.2 [22]. In addition to CD8⁺ and CD4⁺ T cells, TIM-3 is also expressed by regulatory T cells (Tregs) [23], Th17 cells [24, 25], dendritic cells (DCs) [16, 26], myeloid cells [26], natural killer (NK) cells [27], macrophages [28, 29], mast cells [30], and cancer cell subsets (e.g., melanoma, glioma, osteosarcoma, cervical cancer and clear cell renal cell carcinoma) [31–35].

2.1. TIM-3 ligands

Four ligands bind TIM-3 to regulate anti-tumor immunity: Galectin-9 (Gal-9) [36], carcinoembryonic antigen-related cell adhesion molecule 1 (CEACAM1) [37], phosphatidylserine (PtdSer) [38], and high-mobility group box 1 (HMGB1) [16]. Gal-9 is a type C lectin, and was the first identified TIM-3 ligand [39]. This lectin is expressed by myeloid cells, T cells, B cells, mast cells, endothelial cells, tumor cells, and stromal cells and is secreted into the extracellular space [40]. Gal-9 binds to carbohydrates on the immunoglobulin V domain of TIM-3 in mice to induce apoptosis of TIM-3⁺ T cells [36]. The specific binding site of Gal-9 to TIM-3 remains unknown. Gal-9 binding induces the release of HLA-B-associated transcript 3 (BAT3) from the TIM-3 tail, resulting in T cell inhibitory signals and T cell dysfunction [15].

Multiple lines of evidence support the inhibitory effects of the TIM-3/Gal-9 pathway in cancer immunology. CT26 mouse colon tumor cells secrete Gal-9, which causes apoptosis of tumor-infiltrating TIM-3⁺CD8⁺ T cells; in contrast, TIM-3/Gal-9 blockade reduces apoptosis and inhibits tumor growth in mice [41]. In patients with hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC), high numbers of HCC-infiltrating Gal-9⁺ Kupffer cells and TIM-3⁺ T cells appear to correlate with poor prognosis [42]. Transgenic overexpression of either TIM-3 or Gal-9 implicates the TIM-3/Gal-9 pathway in regulating Th1-type immunity by directly triggering Th1 cell death and expanding the suppressive CD11b⁺Ly-6G⁺ cells, which exhibit a phenotype consistent with granulocytic myeloid-derived suppressor cells (MDSCs) [43]. Notably, in human acute myeloid leukemia (AML), a TIM-3/Gal-9 autocrine stimulatory loop drives self-renewal of myeloid leukemia stem cells and leukemic progression through nuclear factor kappa B (NF-κB) and β-catenin signaling. Accordingly, applying anti-human Gal-9 antibodies in mice inhibits xenogeneic reconstitution of human AML [44]. Gal-9 also exerts TIM-3–independent immunological effects. Gal-9 induces pro-inflammatory cytokines from Th1/2 cells and suppresses Th17 cell development in the absence of TIM-3 [45, 46]. In addition, Gal-9 interacts with CD44 to promote expression and stability of forkhead box protein 3 (Foxp3) in induced Treg (iTreg) cells, enhancing their suppressive function [47].

PtdSer is a phospholipid normally localized in the inner leaflet of the plasma membrane that is exposed to the outer leaflet when a cell undergoes apoptosis, stress, or cell activation. PtdSer binds TIM-3 expressed on DCs or macrophages, resulting in phagocytosis of apoptotic cells and cross-presentation [38]. The binding site for PtdSer is located between the FG’ and CC’ loops of the TIM-3 IgV domain, distinct from the Gal-9-binding site on the opposite side of the IgV domain [48]. PtdSer binds to TIM-1 and TIM-4 with higher affinity than TIM-3 [48].

HMGB1 is an alarmin that interacts with multiple ligands such as receptor for advanced glycation endproducts (RAGE), toll-like receptor 4 (TLR4), and interleukin 1 (IL-1) receptor [49]. HMGB1 acts as a universal sentinel for nucleic acids and is required for nucleic acid-induction of innate responses mediated by pattern recognition receptors (PRRs) [50]. TIM-3 expressed by tumor-associated DCs binds HMGB1, which impairs the recruitment of nucleic acids into endosomes and disrupts PRR-mediated innate immune responses [16]. In mouse tumor models, TIM-3 blockade increases the therapeutic efficacy of DNA vaccination and chemotherapy in part by augmenting HMGB1-mediated nucleic acid-sensing systems [16]. These observations suggest that TIM-3 blockade promotes the capability of tumor-derived nucleic acids together with danger signals to activate innate immune response through activation of the PRR-mediated pathway.

CEACAM1 is co-expressed with TIM-3 on activated T cells. CEACAM1 interacts with TIM3 in cis and trans to facilitate cell-surface expression of TIM-3 and TIM-3 inhibitory function [37, 51]. CEACAM1 shares similar structures of IgV domains with TIM-3 and interacts with TIM-3 along the CC’ and FG cleft [37]. CEACAM1 is also expressed by DCs [52], monocytes [53], macrophages [54], and tumor cells (e.g., melanoma cells) [31]. Both CEACAM1 and Gal-9 induce phosphorylation of Tyr256 and Tyr263 in the TIM-3 cytoplasmic tail to promote the inhibition of T-cell receptor (TCR) signaling, although CEACAM1 and Gal-9 bind to different regions of the IgV domain of TIM-3 [15, 37].

Notably, blocking anti-TIM-3 antibodies evaluated to date appear to disrupt the interaction of TIM-3 with PtdSer or CEACAM1 but not HMGB1 [55]. Further investigations of TIM-3–ligand interactions and intracellular downstream signaling are required to obtain a deeper understanding of the regulatory functions of TIM-3.

2.2. Structure and signaling of TIM-3

TIM-3 belongs to the immunoglobulin superfamily and consists of a membrane-distal N-terminal immunoglobulin variable (IgV) domain, a membrane-proximal mucin domain, a transmembrane region, and a cytoplasmic tail [22]. The mouse TIM-3 IgV domain contains two N-linked glycosylation sites, which bind to Gal-9 [36]. The crystal structure of TIM-3 shows that the IgV domain possesses two disulfide bonds that bring the CC’ loop close to the FG loop, resulting in a unique Gal-9-independent ligand-binding cleft that binds CEACAM1 and PtdSer [38, 51]. TIM-3 does not contain classical inhibitory immunoreceptor tyrosine-based inhibition motifs (ITIMs) or immunoreceptor tyrosine-based switch signaling motifs found in other IRs (e.g., PD-1 and TIGIT). The intracellular mechanisms of TIM-3 signaling are not fully understood. Mouse and human TIM-3 cytoplasmic tails contain five conserved tyrosine residues, which can be phosphorylated by src kinases [56] or inducible T cell kinase [57]. Tyr256 and Tyr263 interact with BAT3 and tyrosine kinase FYN [15, 56, 57].

In addition to its ability to induce apoptosis of Th1 cells [36] and act as a coinhibitory receptor, TIM-3 may act as a positive regulator of T cell function [26, 56, 58]. Ectopic expression of TIM-3 in Jurkat T cells leads to the enhancement of TCR-dependent signaling pathways [56]. In addition, CD8⁺ T cells from Tim-3^−/− mice have impaired recall responses upon infection with the acute armstrong strain of lymphocytic choriomeningitis virus, and enforced expression of TIM-3 in vivo drives the generation of short-lived effector CD8⁺ T cells in an AKT/mammalian target of rapamycin (mTOR)-dependent manner [58]. In this experimental model, sustained positive signals via TIM-3 eventually promote the exhaustion of TIM-3⁺ T cells by enhancing initial T cell activation and the generation of short-lived effector cells at the expense of memory T cell generation [58]. To explain these conflicting findings, Kuchroo and colleagues suggested that BAT3 acts as a master regulator of TIM-3 signaling. BAT3, which is a repressor of TIM-3 inhibition, binds to the TIM-3 cytoplasmic tail, recruiting the catalytically active form of LCK to permit TCR stimulation. However, when TIM-3 binds to its ligands CEACAM1 [37] and Gal-9 [57], Tyr256 and Tyr263 are phosphorylated, and BAT3 is released from the TIM-3 tail, allowing TIM-3 to exert inhibitory functions [15]. In the absence of BAT3, T cells lose the ability to produce large amounts of IFN-γ and IL-2 [15]. Notably, FYN and BAT3 bind to the same region of TIM-3, and FYN can promote T cell anergy through the activation of glycosphingolipid-enriched microdomains 1 [59]. Therefore, the potential competition between FYN and BAT3 for binding to TIM-3 may contribute to the inhibition of TCR signaling [60] (Fig. 1).

Figure 1. Structure and signaling of TIM-3.

1) In the absence of TIM-3 ligands, BAT3 binds the TIM-3 cytoplasmic tail and recruits the active form of LCK to the TCR complex to activate TCR signaling. 2) In the presence of ligand, BAT3 is released from the TIM-3 cytoplasmic tail, resulting in the recruitment of phosphatases, dephosphorylation of LCK, and inhibition of TCR signaling. Tyrosine kinase FYN can be recruited to the TIM-3 tail after the release of BAT3. 3) TIM-3 blockade may promote TIM-3-BAT3 interaction and TCR signaling.

2.3. TIM-3 in cancer immunology

TIM-3 was initially associated with negative regulation of T cell functions in autoimmune diseases (experimental autoimmune encephalomyelitis, type 1 diabetes and multiple sclerosis) [21, 36, 61, 62]. High levels of TIM-3 expression by PD-1⁺ T cells correlate with T cell dysfunction and dampen T cell response during chronic viral infection and cancer [6, 63–69]. In patients with advanced melanoma, TIM-3 is co-expressed by a fraction of PD-1⁺ tumor antigen-specific CD8⁺ T cells in the periphery and at tumor sites, which are more dysfunctional than PD-1⁺TIM-3⁻ and PD-1⁻TIM-3⁻ CD8⁺ T cells [6]. Dual PD-1/TIM-3 blockade enhances cytokine production and proliferation of human antigen-specific CD8⁺ T cells [6]. There have been similar findings in mouse tumor-bearing models, and targeting TIM-3 and PD-1 pathways with monoclonal antibodies impede tumor growth and prolong survival in vivo [69, 70]. In addition, TIM-3 overexpression in T cells exacerbates tumor growth in the EL4 mouse model of lymphoma [43]. Increased TIM-3 expression on CD8⁺ tumor-infiltrating lymphocytes (TILs) correlates with poor prognosis in many solid tumors, including non-small cell lung cancer (NSCLC), renal cancer, HBV-associated HCC, and prostate cancer [42, 71–73]. TIM-3 upregulation by CD8⁺ TILs occurs upon adaptive resistance to PD-1 inhibitors in mice and humans with lung cancer, and TIM-3 blockade confers clinical benefits to PD-1 refractory mice [74]. Interestingly, TIM-3 and PD-1 are not only upregulated by tumor-antigen specific dysfunctional CD8⁺ T cells, but also activated vaccine-induced T cells in patients with advanced melanoma. Dual blockade of PD-1 and TIM-3 enhances the expansion and cytokine production of vaccine-induced CD8⁺ T cells upon stimulation with cognate antigen in vitro [68].

TIM-3 is upregulated on Tregs, which are potent suppressors of antitumor immune responses in the TME [75]. TIM-3⁺ Tregs appear to represent a subset of activated Tregs [75, 76] present in many solid tumors, including NSCLC [23], which correlate with tumor burden and progression. In healthy donors and patients with cancer, most peripheral NK cells express TIM-3. TIM-3 is expressed by all mature CD56^dimCD16⁺ NK cells and is expressed heterogeneously in the immature CD56^brightCD16⁻ NK-cell subset in blood from healthy adults [27]. TIM-3 blockade reinvigorates peripheral NK cell function in vitro in the presence of cognate ligands [77]. In contrast to T cells, circulating TIM-3⁺ NK cells in healthy donors are not dysfunctional [27]. Compared with healthy donors, peripheral NK cells in patients with melanoma express significantly higher levels of surface TIM-3, although this increase is modest [77]. Higher TIM-3 expression on NK cells is associated with advanced stages of melanoma and poor prognostic clinical parameters. Blocking TIM-3 partially increases cytotoxicity and IFN-γ production by circulating NK cells in melanoma patients [77]. Notably and in sharp contrast with CD8⁺ TILs, tumor-infiltrating NKs are dysfunctional and do not upregulate TIM-3 compared to circulating NK cells in advanced melanoma [78]. In addition, dysfunctional NK cells in metastatic tumors downregulate many inhibitory and costimulatory receptors [79–84].

TIM-3 is constitutively expressed by APCs and is further upregulated by DCs in the TME [16]. In a mouse model of breast cancer, intratumoral CD103⁺ DCs upregulate TIM-3 expression and anti-TIM-3 antibody promotes chemokine (C-X-C motif) ligand 9 expression by cross-presenting BATF3⁺CD103⁺ DCs (type 1 conventional DCs) to enhance the CD8⁺ T cell-dependent response to the chemotherapeutic paclitaxel [85]. In this model, the effects of anti-TIM-3 antibody appear to be Gal-9-dependent. The precise mechanism used by anti-TIM-3 antibody to reprogram DCs in the TME and relevance of these data to humans remain unknown.

Transgenic overexpression of TIM-3 on T cells suppresses adaptive immunity in a murine model of EL4 thymoma by promoting the generation of MDSCs in a TIM-3/Gal-9–dependent manner [43]. Interestingly, one study recently reported the accumulation of lymphoid cells and monocytic MDSCs in PD-1 refractory patients with NSCLC [86]. Gal-9 but not PD-L1 expression is negatively correlated with IFN-γ production by peripheral blood mononuclear cells (PBMCs) upon stimulation with anti-CD3 antibodies. Dual PD-1/TIM-3 blockade restores cytokine production by PBMCs in vitro. Together, these findings suggest that TIM-3 and Gal-9 may be implicated in resistance to PD-1 blockade both in patients with primary or acquired secondary resistance to anti-PD-1.

Soluble TIM-3 is released upon shedding by sheddases (ADAM10 and ADAM17) [87] and may impair antitumor immunity. Highly levels of elevated soluble TIM-3 correlate with high risk of HCC and poor survival of patients with HCC [88], and soluble TIM-3 enhances B16F1 tumor growth in mice by inhibiting T cell-mediated antitumor immunity and antigen-specific stimulation [89]. TIM-3 is also expressed by tumor cells in melanoma, glioma, osteosarcoma, cervical cancer, and clear cell renal cell carcinoma and appears to facilitate their migration, proliferation, and invasion [31–35]. TIM-3 is highly expressed by leukemic stem cells (LSCs) in AML but not in normal hematopoietic stem cells [90–94], and represents a potential therapeutic target. In a mouse xenograft model, anti-TIM-3 blockade with Fc-binding capacity eradicates human AML LSCs without affecting normal human hematopoiesis [90, 94]. The mechanisms supporting the role of tumor-expressed TIM-3 in the regulation of antitumor immune responses remain unknown. Finally, a meta-analysis of published studies totaling 869 patients with advanced solid tumors suggests that high TIM-3 expression in the TME is associated with advanced tumor stage and shorter overall survival, supporting the role of TIM-3 as an inhibitory receptor in antitumor immune responses [95].

2.4. Genetic and epigenetic regulation of TIM-3 in cancer

Human TIM-3 contains a large number of single nucleotide polymorphisms (SNPs) [96]. These SNPs are associated with TIM-3 expression and risk of cancer including gastrointestinal tract cancer [97], breast cancer [98], renal cell carcinoma [99], HBV-related HCC [73], pancreatic cancer [100], NSCLC, and non-Hodgkin lymphoma [101]. In addition, germline mutations c.245A>G (p.Tyr82Cys) and c.291A>G (p.Ile97Met) in the TIM-3 IgV domain result in misfolding and post-translational modifications of TIM-3, abrogating TIM-3 cell surface expression by T cells and myeloid cells. These mutations were found in 16 of 27 patients with a rare form of non-Hodgkin T cell lymphoma, called subcutaneous panniculitis-like T cell lymphoma (SPTCL), which is often associated with hemophagocytic lymphohistiocytosis (HLH). Misfolding of TIM-3 promotes persistent immune activation and increases the production of inflammatory cytokines, including tumor necrosis factor-α (TNF-α) and IL-1β, promoting HLH and SPTCL. Such findings support the inhibitory function of TIM-3 [102–104].

Tumor-secreted molecules such as PGE₂, which activate protein kinase A (PKA), upregulate transcriptomic TIM-3 expression in T cells [105]. CpG islands in the promoter regions of TIM-3 appear to be significantly hypomethylated in breast tumor tissues compared with normal tissues [106]. Also, bindings of repressive histones H3K9me3 and H3K27me3 in the TIM-3 promoter are significantly reduced in tumor tissues compared with normal tissues [106, 107].

2.5. Targeting TIM-3 in cancer immunotherapy

Ample evidence in preclinical models supports dual PD-1/TIM-3 blockade to augment the tumor antigen-specific T cell response and promote tumor regression. Based on these findings, several TIM-3-targeting antibodies are being tested in clinical trials (Table 1). TSR-022 is a humanized anti-TIM-3 IgG4 antibody developed by Tesaro/GlaxoSmithKline that is being evaluated alone and in combination with anti-PD-1 (TSR-042) and anti-lymphocyte activation gene-3 (LAG-3) (TSR-033) in an ongoing phase I/II study for PD-1 refractory melanoma and NSCLC patients. Anti-TIM-3 antibody alone or in combination with anti-PD-1 is safe, and the toxicity profile is similar to PD-1 inhibitors. Clinical benefits have been observed in the high dose of TSR-022 combination group with a 15% overall response rate (ORR) (3/20) and 40% stable disease (8/20) [108]. In a phase Ib clinical trial with anti-TIM-3 (MGB453) as monotherapy and in combination with anti-PD-1 (PDR001) in AML and high-risk myelodysplastic syndrome, 8 of 16 (50%) patients with HR-MDR and 4 of 14 (29%) patients with newly diagnosed AML exhibited clinical responses [109].

Table 1:

Active clinical trials targeting TIM-3, TIGIT, or CD112R

Target	Agents (manufacturer)	Protocol (tumor types)	Combination agents
TIM-3	TSR-022, humanized IgG4 (Tesaro)	NCT02817633, phase I (advanced solid tumors)	TSR-023 (anti-PD-1), TSR-033 (anti-LAG-3)
		NCT03580508, phase II (advanced, localized/unresectable HCC)	TSR-023 (anti-PD-1)
		NCT03307785, phase I (advanced solid tumors)	Niraparib, TSR-042 (anti-PD-1), Bevacizumab (anti-VEGF), platinum-based chemotherapy
	MBG453, humanized IgG4 (Novartis)	NCT02608268, phase I/II (advanced malignancies)	PDR001 (anti-TJ-1)
		NCT03066648, phase I (AML or high risk MDS)	Decitabine, PDR001 (anti- PD-1)
		NCT03940352, phase I (AML or high risk MDS)	HDM201, Venetoclax
		NCT03961971, phase I (GBM)	Spartalizumab (anti-PD-1)
	Sym023, human IgG1 (Symphogen)	NCT03489343, phase I (advanced malignancies)	None
		NCT03311412, phase I (advanced malignancies)	Sym021 (anti-PD-1), Sym022 (anti-LAG-3)
	LY3321367, human IgG1 (Eli Lilly and Company)	NCT03099109, phase I (advanced relapsed/refractory solid tumors)	LY33000054 (anti-PD-L1)
	BGB-A425, humanized IgG1 (BeiGene)	NCT03744468, phase I/II (locally advance solid tumors)	Tislelizumab (anti-PD-1)
	RO7121661, bispecific anti-PD-1/TIM-3 (Hoffmann-La Roche)	NCT03099109, phase I (advanced solid tumors)	None
	INCAGN02390, IgG1k (Incyte Corporation)	NCT03652077, phase I (advanced malignances)	None
		NCT04370704, phase I/II (advanced malignances)	INCMGA00012 (anti-PD-1), INCAGN02385 (anti-LAG-3)
	BMS-986258, (Bristol-Myers Squibb)	NCT03446040, phase I (advanced cancer)	Nivolumab (anti-PD-1), rHuPH20
	SHR-1702 (Jiangsu HengRui)	NCT03708328, phase I (advanced malignances)	Camrelizumab (anti-PD-1)
TIGIT	BMS-986207, human IgG1 (Bristol Myers Squibb)	NCT04150965, phase I/II (multiple myeloma with relapse)	None
	BGB-A1217, humanized IgG1 (BeiGene)	NCT04047862, phase I/Ib (metastatic solid tumors)	Tislelizumab
	Tiragolumab, MTIG7192A, human IgG1 (Genentech)	NCT02794571, phase Ia/b (advanced tumors)	Atezolizumab (anti-PD-L1), carboplatin, cisplatin, pemetrexed, paclitaxel
		NCT04045028, phase I (myeloma and lymphoma)	Rituxomab (anti-CD20), Daratumumab/rHuPH20 (anti- CD38/hyaluronidase PH20)
		NCT04045028, phase Ib/II (urothelial carcinoma)	Atezolizumab (anti-PD-L1)
		NCT03281369, phase Ib/II (gastroesophageal cancers)	Atezolizumab (anti-PD-L1)
		NCT03193190, phase Ib/II (pancreatic adenocarcinoma)	Atezolizumab (anti-PD-L1)
		NCT03563716, phase II (lung cancer),	Atezolizumab (anti-PD-L1)
		NCT04294810, phase III (PD-L1 positive non-small lung cancer)	Atezolizumab (anti-PD-L1)
		NCT04300647, phase II (metastatic or recurrent PD-L1 positive cervical cancer)	Atezolizumab (anti-PD-L1)
		NCT04256421, phase III (small cell lung cancer)	Atezolizumab (anti-PD-L1), carboplatin, etoposide
	AB154, humanized IgG1 (Arcus Biosciences)	NCT03628677, phase I (advanced solid malignancies),	Zimberelimab (anti-PD-1)
		NCT04262856, phase II (PD-L1 positive non-small lung cancer)	Zimberelimab (anti-PD-1), AB928 (anti-A2a/bR antagonist)
	ASP8374, human IgG4 (Astella Pharma Global Development)	NCT03260322, phase Ib (advanced tumors)	Pembrolizumab (anti-PD-1)
	MK-7684, humanied IgG1 (Merck Sharp & Dohme)	NCT4305041, phase I/II (melanoma)	Pembrolizumab (anti-PD-1), MK-1308 (anti-CTLA-4)
		NCT04303169, phase I/II (PD-1 refractory melanoma)	Pembrolizumab (anti-PD-1)
	M6223, TIGIT inhibitor (Merck Sharp & Dohme)	NCT04457778, phase I (metastatic solid tumors)	Bintrafusp alfa (anti-PD-L1/TGFβ)
	C0M902, TIGIT inhibitor (Compugen)	NCT04354246, phase I (advanced malignancies)	None
CD112R	COM701, inhibitor (Compugen)	NCT03667716, phase I (advanced solid tumors)	Nivolumab (anti-PD-1)

Antibodies targeting TIM-3 and TIGIT and drugs targeting CD112R found on ClinicalTrials.gov (as of August 2020) that are active in clinical trials for the indicated tumor types and with the combinatorial treatment listed. AML, acute myeloid leukemia; GBM, glioblastoma multiforme; HCC, hepatocellular carcinoma; Ig, immunoglobulin; LAG-3, lymphocyte-activation gene 3 protein; mAb, monoclonal antibody; MDS: myelodysplastic syndrome; PD-1, programmed cell death 1; PD-L1, programmed death ligand 1

3. TIGIT

TIGIT (also called WUCAM, Vstm3, VSIG9) is an IR of the Ig superfamily expressed by multiple immune cells. TIGIT is part of a complex regulatory network involving multiple ligands [e.g., CD155 (PVR/NECL-5), CD112 (nectin-2/PVRL2)], two competitive IRs, CD96/TACTILE and CD112R/PVRIG, and one known competing co-stimulatory receptor, DNAM-1/CD226 [17, 18, 20, 110, 111]. This pathway is reminiscent of the CD28/CTLA-4/CD80/CD86 pathway, for which IRs and costimulatory receptors compete for binding to the same ligands [112]. Tigit^−/− mice do not develop autoimmunity, in contrast with Ctla4^−/− mice [113]. However, Tigit^−/− mice develop more severe experimental autoimmune encephalitis than wild-type (WT) mice when immunized with myelin oligodendrocyte glycoprotein [113]. In humans, TIGIT is expressed on activated T cells, follicular helper T cells (Tfh), Tregs and NK cells in PBMCs [17, 114–116]. In the TME, TIGIT is upregulated by Tregs and TILs [118, 119]. DNA hypomethylation and binding of repressive histones (H3K9me3 and H3K27me3) in the TIGIT promoter appear to be significantly reduced in tumor tissues and compared with normal tissues, contributing to TIGIT upregulation [106]. In both mice and humans, TIGIT is co-expressed with PD-1 on CD8⁺ TILs and on tumor antigen-specific CD8⁺ T cells [120, 121]. TIGIT is also co-expressed with IRs such as TIM-3 and LAG-3 on exhausted CD8⁺ T cell subsets in the TME [120, 121]. TIGIT expression delineates Tregs from activated effector CD4⁺ T cells and its upregulation is associated with hypomethylation and Foxp3 binding at the Tigit locus [122]. Unlike mouse NK cells, TIGIT is highly expressed on human NK cells and regulates their antitumor activity [123]. Although TIGIT⁺ NK cells are more lytic and more mature than TIGIT⁻ NK cells, they display lower cytotoxicity against CD155⁺ MHC class I-deficient melanoma [124]. In metastatic tumors, NK cells are found at a low frequency and they downregulate both TIGIT and CD226 expression in contrast with CD8⁺ T cells [124]. Membrane-bound CD155 promotes CD226 internalization and degradation, impeding NK cell-mediated tumor reactivity [124].

3.1. TIGIT ligands

TIGIT binds CD155 and CD112, which are members of the nectin protein family. TIGIT binds CD155 with high affinity as compared with its competitors and binds CD112 with low affinity [18] (Table 2). CD155 and CD112 are expressed by circulating and APCs including DCs, as well as many non-hematopoietic cells including tumor cells of different histological types [18, 120, 125–127]. CD155 and CD112 bind other nectin-like proteins including CD113 [18]. CD155 expression is upregulated upon reactive oxygen species-dependent activation of the DNA damage response, which regulates interactions among MDSCs, T cells, and NK cells [128, 129]. Fap2 protein from Fusobacterium nucleatum, an anaerobic Gram⁻ commensal bacteria associated with colorectal carcinoma, binds TIGIT to inhibit NK- and T-cell mediated tumor reactivity. Such findings support the role of the gut microbiome in regulating antitumor immune responses via TIGIT [130].

Table 2:

Ligand binding affinities for human CD226, TIGIT, CD96, and CD112R

Ligand/receptor affinity	CD226	TIGIT	CD96	CD112R
CD155	114–199 nM	1–3 nM	37.6 nM	-
CD112	0.31–8.97 μM	Not measurable	-	88 nM

Ligand-binding affinities for human TIGIT, CD226, CD96, and CD112R as previously reported [17, 18, 20, 110, 111].

3.2. Structure and signaling of TIGIT

TIGIT contains one extracellular Ig variable domain, a transmembrane domain, and a cytoplasmic tail containing two inhibitory motifs conserved in mice and humans: an ITIM and an Ig tail-tyrosine (ITT)-like motif [17, 18, 113, 114] (Fig. 2). Upon TIGIT/CD155 interaction, phosphorylation of the ITT-like motif (Tyr225) triggers TIGIT inhibitory signals in humans, whereas both the ITT and the ITIM motifs induce inhibitory signals in mice [17, 131]. In NK cells, TIGIT phosphorylation recruits the cytosolic adapters Grb2 [131] and β-arrestin 2 [132] to bind SH2-containing inositol phosphatase-1 (SHIP1). SHIP1 inhibits phosphoinositide 3 kinase and mitogen-activated protein kinase signaling [131]. SHIP1 also impedes the activation of TNF receptor-associated factor 6 (TRAF6) and NF-κB and disrupts IFN-γ production by NK cells [132].

Figure 2: TIGIT/CD96/CD112R inhibitory pathways.

1) CD155 engagement with TIGIT drives DCs into producing IL-10 instead of IL-12. 2) TIGIT competes with CD226, binds CD226 in cis to reduce its binding stability to CD155, and induces direct intrinsic inhibitory signal via Grb2, SHIP1, and β-arrestin recruitment at the ITT-like motif. 3) TIGIT engagement promotes Treg stability, Foxp3 binding to the TIGIT locus, and suppressive functions with the production of Fgl2 and IL-10. 4) Fap2 protein from F. nucleatum can bind TIGIT, triggering direct inhibitory signals in the TME. 5) CD96 competition with CD226 for CD155 dampens stimulatory signals and its ITIM binds SHIP1/2 adaptors to trigger inhibitory signals while its YXXM might bind SH2 to promote stimulatory signals in T cells. 6) CD112R competition with CD226 for CD112 dampens stimulatory signals and its ITIM binds SHP1/2 adaptors to trigger inhibitory signals.

3.3. TIGIT in cancer immunology

TIGIT inhibits adaptive and innate immunity through multiple mechanisms [10, 11, 17, 114] (Fig. 2). First, in a mouse model, TIGIT binds CD155 on DCs to impair T cell functions [18]. TIGIT induces CD155 phosphorylation and triggers a signaling cascade promoting tolerogenic DCs that produce less IL-12 and more IL-10 [18]. Second, TIGIT exerts direct T cell-intrinsic inhibitory effects by dampening TCR signaling, leading to decreased T cell proliferation and function [113, 115, 126]. TIGIT also inhibits mouse and human NK cell degranulation, cytokine production, and NK cell-mediated cytotoxicity of CD155-expressing tumor cells [17, 131–133]. Notably, CD155⁺ MDSCs act on TIGIT⁺ NK cells to decrease phosphorylation of ZAP70/Syk and ERK1/2 and reduce their cytolytic capacity [129].

Third, TIGIT disrupts CD155-mediated DNAM-1/CD226 activation. CD226 is a costimulatory receptor widely expressed by immune cells including T cells, NK cells, monocytes, and platelets [134, 135]. This receptor is directly involved in tumor recognition by T cells and NK cells in mice and humans [134, 136]. CD226-deficient mouse CD8⁺ T cells and NK cells display immunological synapse defects impairing antitumor immunity [137, 138]. CD226 also associates with lymphocyte function-associated antigen 1 (LFA-1) to enhance TCR signaling [139]. TIGIT impedes CD226-mediated activation because it binds CD155 with higher affinity than CD226 [17, 18, 113], and directly binds CD226 in cis, impairing its homodimerization and binding capacity to CD155 [121].

Fourth, multiple lines of evidence suggest that the balance of TIGIT/CD226 expression plays a central role in regulating the effector function of T cells and NK cells. TIGIT knockdown with shRNA in TCR-activated CD4⁺ T cells increases T-bet expression and IFN-γ production, which are abolished upon CD226 or CD155 blockade. In contrast, CD226 silencing decreases T-bet expression and IFN-γ production [126]. Blocking anti-CD226 antibodies abrogates the immunological effects of dual PD-1 and TIGIT blockade on CD8⁺ T cell-mediated tumor in vitro and in CT26 tumor-bearing mice [120, 121]. Interestingly, anti-PD-1 blocking mAb and agonistic anti-GITR mAbs increase overall survival of mice bearing MC38 colon adenocarcinoma tumors. Notably, in MC38 tumor-bearing mice treated with anti-PD-1 blocking mAbs and agonistic anti-GITR mAbs, PD-1 blockade counteracts CD8⁺ T cell dysfunction by inhibiting Scr homology domain 2-containing tyrosine phosphatase (SHP2)-mediated CD226 dephosphorylation, while anti-GITR mAbs decrease TIGIT expression [140]. Such findings suggest that in addition to PD-1 and TIGIT blockades, other ICBs can exert beneficial clinical effects by favorably tipping the balance between CD226 and TIGIT in CD8⁺ T cells.

Fifth, TIGIT acts in Tregs to augment immunosuppressive function and stability. TIGIT is highly expressed by a subset of natural Tregs in mice [118] and the majority of Tregs in humans [118, 119, 141]. In Tregs, the TIGIT locus is hypomethylated and includes Foxp3 binding sites [122]. TIGIT⁺ Tregs are more suppressive than TIGIT⁻ Tregs in healthy donors and patients with melanoma [119, 141]. Further, TIGIT⁺ Tregs in the periphery and at tumor sites upregulate many Treg gene signature markers as compared with TIGIT⁻ Tregs [118] including FOXP3, IKZF2, NRP-1, CTLA-4, PD-1, and LAG-3 [118, 119]. TIGIT⁺ Tregs also suppress proinflammatory Th1 and Th17 but not Th2-type T cell responses [118, 142]. Upon TIGIT ligation, TIGIT⁺ Tregs produce IL-10 and fibrinogen-like protein 2 (FGL2), which mediate T cell suppression [118]. Compared to the periphery, Tregs exhibit lower CD226 expression in metastatic melanoma, resulting in an increased TIGIT/CD226 ratio [119]. TIGIT and CD226 oppose each other to augment or disrupt Treg suppression and stability, respectively [119]. A high TIGIT/CD226 ratio in Tregs appears to correlate with increased Treg frequencies in tumors and poor clinical outcome upon ICB. Whether the TIGIT/CD226 ratio in Tregs represent a biomarker of clinical response to ICB in patients with solid tumors remains to be determined.

3.4. Targeting TIGIT in cancer immunotherapy

Accumulating preclinical data suggest that dual PD-1/TIGIT blockade is a promising combinatorial cancer immunotherapy. While either single blockade does not significantly impede growth of CT26 tumors in mice, dual TIGIT and PD-1/PD-L1 blockade synergizes to augment CD8⁺ T cell-mediated tumor reactivity, resulting in protective memory T cells, complete tumor rejection, and prolonged overall survival [121, 143, 144]. Dual PD-1/TIGIT blockade in patients with melanoma also augments functions of isolated tumor-antigen-specific CD8⁺ T cells and TILs compared with effects of single blockade [120, 145]. Further, dual PD-L1/TIGIT blockade (atezolizumab [atezo]/tiragolumab [tira]) provides superior clinical benefits with improved ORR and progression-fee survival compared with PD-L1 blockade alone as a first-line therapy for patients with PD-L1-positive NSCLC, despite similar toxicity profiles [146]. This study included 135 NSCLC patients with an ORR of 37% for atezo/tira compared with 21% for atezo and placebo. In PD-L1⁺ tumors (tumor proportion score > 50%), ORR was 66% for atezo/tira vs 24% for atezo and placebo. Combinatorial therapy with anti-PD1 and anti-TIGIT antibodies is being evaluated in an ongoing phase III study.

The effects of dual PD-1/TIGIT blockade in mouse tumor models and in vitro are abrogated upon CD226 blockade, suggesting that TIGIT blockade acts primarily by tipping CD155-mediated signaling toward CD226 activation [120, 121]. In addition, PD-1 induces SHP2-mediated CD226 dephosphorylation, supporting the additive effects of dual PD-1/TIGIT blockade in enhancing CD226 signaling [140]. Membrane-bound CD155 plays a critical role in mediating CD226 downregulation by immune cells in the TME via CD226 internalization and degradation, supporting the role of CD155-mediated immune dysfunction [124]. CD226 downregulation by CD8⁺ TILs, which has been observed in many solid tumors, including melanoma, may limit the effects of dual PD-1/TIGIT blockade in patients with cancer [120, 147].

TIGIT blockade or TIGIT deletion augments NK cell-mediated antitumor reactivity in vitro and in vivo [17, 129, 131, 133, 148]. One study of B16 melanoma and CT26 lung metastasis mouse models reported that TIGIT blockade alone or in combination with PD-1 blockade acted primarily on NK cells to augment CD8⁺ T cell-mediated antitumor responses and impede tumor growth. In the same study, NK cell-specific TIGIT deficiency and NK cell depletion abrogated the effects of TIGIT blockade [143]. These findings do not fit well with the many observations supporting that TIGIT blockade alone fails to significantly augment CD8⁺ T cell immunity and promote tumor rejection in WT mice transplanted with solid tumors [121, 149] and in expanding tumor antigen-specific CD8⁺ T cells [120] compared with dual PD-1/TIGIT blockade. Mechanisms supporting potential helper effects of NK cells on CD8⁺ TILs as well as the relevance of these findings to cancer patients remain unclear. Whether and how NK cells contribute to environmental cues guiding CD8⁺ T cell priming, maturation, and memory differentiation needs to be determined. Interestingly, IL-15 together with TIGIT blockade increases NK cell-mediated melanoma cytotoxicity in vitro and decreases tumor metastasis in mouse melanoma models [124]. These findings support the development of novel combinatorial immunotherapy with IL-15 and TIGIT blockade to promote NK cell-mediated destruction of MHC class I-deficient melanoma, which is refractory to CD8⁺ T cell-mediated immunity.

One study in mouse tumor models showed that ICB with Fc variants of anti-TIGIT mAbs with selective FcγR co-engagement on APCs enhances antigen-specific T cell responses and tumor reactivity without evidence of Treg depletion [150]. Whether the antitumor effects of anti-TIGIT antibodies in cancer patients are Fc-dependent will need to be further investigated in ongoing clinical trials with different Fc-engineered anti-TIGIT mAbs (Table 1).

4. CD96

CD96 (also known as TACTILE) is a receptor of the Ig superfamily, first described as an activation receptor [151]. In humans, CD96 is expressed on T cells, NK cells, and gammadelta T cells [151, 152]. CD96 is a marker of T cell activation, which is co-expressed with CD226, TIGIT, and PD-1 on T cells upon TCR activation [152]. In multiple solid tumors, increased CD96 mRNA expression appears to correlate with increased expression of T cell markers, and the expression of CD96 and CD8A are higher in metastatic tumors than in primary tumors [152]. CD96 is also expressed by acute lymphoblastic leukemia, in AML and is a leukemic stem cell-specific marker in human acute myeloid leukemia [153].

4.1. CD96 ligands

CD96 competes with CD226 and TIGIT for binding to CD155 [17, 18, 110, 111]. The binding affinity of CD96 to CD155 is between that of TIGIT (high affinity) and DNAM-1 (low affinity) (Table 2). Notably, mouse CD96 also binds CD111 [18, 154].

4.2. Structure and signaling of CD96

CD96 contains three extracellular Ig-like domains, a transmembrane sequence, and a short intracellular domain. The first Ig domain of CD96 is necessary and sufficient for nectin-like protein 5 (NECL-5) binding via the canonical “lock-and-key” docking mode, although the second and third Ig domains of CD96 can influence the interaction [155]. In humans, alternative splicing of exon 4 generates two variants containing different second external Ig domains with either an I-like (variant 1) or a V-like (variant 2) Ig domain [156]. Variant 2 is more abundant than variant 1, and exhibits higher binding capacity to CD155, which promotes cell-to-cell adhesion [125, 156]. Both mouse and human cytoplasmic tails of CD96 contain a short basic/proline rich motif and a single ITIM motif, suggesting possible inhibitory signaling [156]. However, the human CD96 cytoplasmic tail also contains an YXXM motif, which is found in inducible T cell costimulator (ICOS) protein and CD28 and binds to the SH-2 domain [156], suggesting potential inhibitory or stimulatory function.

4.3. CD96 in cancer immunology and immunotherapy.

Whether CD96 acts as a costimulatory or an inhibitory receptor (Fig. 2) is under debate. CD96 was first described as an activating receptor [125]. In that study, the investigators tested the cytotoxicity of human polyclonal NK cell lines and NK92 against P815 cells in the presence of Abs that bound to the Fc receptor on P815 and engaged CD96. Anti-CD96 antibody increased the lysis of P815 by human polyclonal NK cell lines, although not as efficiently as the engagement of DNAM-1 [125]. The same anti-CD96 antibody failed to promote tumor lysis in the presence of primary NK cells, and did not influence the killing of CD155-expressing ovarian carcinoma or myeloma cells by NK cells, whereas anti-DNAM-1 significantly blocked NK cell degranulation and cytotoxicity [157, 158]. On one hand, CD96 appears to impede antitumor immune responses in mice. In an experimental mouse lung metastasis model with CD155-expressing tumors, cd96^−/− mice developed less metastases via increased IFN-γ production by NK cells compared to WT mice [159]. CD96 blockade also inhibited experimental tumor metastasis in three different tumor models in an IFN-γ-, CD226- and NK cell-dependent fashion [160]. Blocking CD96 impedes primary tumor growth in many mouse tumor models, in a CD8⁺ T cell and CD226-dependent fashion. CD96 blockade is more effective in dual combination with anti-PD-1, anti-CTLA-4, or anti-TIGIT blockade [161]. Such studies fit well with the observation that PD-1 and CTLA-4 blockade are more effective in the setting in which CD155 is limiting [162]. Anti-CD96 antibody also enhances the clinical effects of dual PD-1/TIGIT blockade.

On the other hand, CD96 appears to exhibit costimulatory function both in mouse and human CD8⁺ T cells [163]. Anti-CD96 mAb co-crosslinking with anti-CD3 mAb promotes T cell proliferation, phosphorylation of the MEK-ERK pathway, expression of markers and transcription factors associated with antigen-specific activation like T-bet and Eomes, and production of proinflammatory cytokines such as IFN-γ and TNF-α. Using OT-I T cell transfer and a syngeneic tumor model, blockade or genetic ablation of CD96 results in functionally antigen-specific CD8⁺ T cells with no increased CD8⁺ T cell infiltrates or controlled tumor growth. The paradoxical observation of CD8⁺ T cell activation without clinical activity in mouse-bearing tumors treated with CD96 blockade remains to be explained. Additional studies are needed to determine whether CD96 blockade regulates CD8⁺ T cell priming, maturation, and memory differentiation in cancer immunotherapy. Also to be determined is whether PD-1 blockade may help enhance the immunological and clinical activity of the anti-CD96 antibody used in this study. Because of the conflicting findings on CD96 blockade reported in different mouse experimental tumor models with different anti-CD96 antibodies, it will be important to compare these antibodies in vitro and in vivo to further understand their mechanisms of action in CD8⁺ T cells and NK cells.

5. CD112R

CD112R is a recently described IR and a member of the Ig superfamily [20]. CD112R is expressed by T cells and NK cells in human PBMCs as well as in numerous solid tumors [164]. CD112R is also a member of the Ig superfamily [20]. This IR competes with CD226 for binding to CD112, which is expressed by DCs and the majority of human tumor cell lines [20]. In mouse, CD112R also binds CD155 with low affinity [165].

5.1. Structure and signaling of CD112R

CD112R contains an extracellular Ig-like domain, a transmembrane sequence, and a cytoplasmic tail that contains two tyrosine phosphorylation sites including an ITIM-like motif [20]. The binding of CD112R to its ligand recruits SHP1 and SHP2 that trigger the inhibition of the TCR/CD28 signaling in T cells [20] (Fig. 2).

5.2. CD112R in cancer immunology and immunotherapy

Several lines of evidence suggest that CD112R blockade combined with TIGIT blockade increases antitumor immune responses in mice and in vitro. In mice transplanted with MC38 tumors, CD112R-deficient mouse CD8⁺ T cells develop a stronger antigen-specific effector response compared with WT CD8⁺ T cells, resulting in reduced tumor growth compared to WT mice, and PD-L1 blockade further enhances the inhibition of tumor growth [165]. Dual CD112R/PD-L1 blockade also reduces tumor growth and increases overall survival [165]. CD112R blockade alone or together with either TIGIT blockade or PD-1 blockade increases cytokine production by TILs from ovarian, endometrial, and lung tumors in the presence of an allogeneic melanoma cell line expressing surface-bound anti-CD3 antibody [164]. The relevance of these findings to human cancer and autologous CD8⁺ T cell responses against well-defined tumor antigens remains to be demonstrated. Finally, CD112R blockade synergizes with TIGIT blockade to enhance human NK cell-triggered antibody-dependent cellular cytotoxicity against breast tumor cell lines in vitro [148]. CD112R blockade is being evaluated in an ongoing clinical trial (NCT03667716) with CD112R inhibitor COM701 (Compugen) alone or in combination with PD-1 blockade (nivolumab) (Table 1).

6. Concluding remarks

TIM-3 is expressed by multiple innate and adaptive immune cells and acts as a negative regulator of antitumor immunity. Ongoing clinical trials with anti-TIM-3 antibodies provide a unique opportunity to determine the immunogenicity and clinical activity of anti-TIM-3 antibodies in patients with cancer. Although TIM-3 binds many ligands, anti-TIM-3 antibodies with proven blocking capacity in vitro appear to bind PtdSer and CEACAM1 but not Gal-9–9. Whether PtdSer, CEACAM1, or TIM-3 expression in the TME may serve as potential biomarkers of response to TIM-3 blockade remains to be determined. Because of the superiority of dual PD-1/TIM-3 blockade in vitro and in experimental tumor models compared with single blockade, it is likely that the PD-1 and TIM-3 pathways are non-redundant. However, the mechanisms of action supporting the activity of dual PD-1/TIM-3 blockade remain to be defined. It will also be essential to identify the TIM-3⁺ cell subsets targeted by TIM-3 blockade that are responsible for the potential immunological and clinical effects in vivo. There is strong evidence in mice and humans that targeting TIGIT together with PD-1 with blocking antibodies augments antitumor immune responses and confers clinical benefits in cancer patients. However, such findings need to be confirmed in large randomized clinical trials. Whether CD112R blockade together with PD-1 blockade or TIGIT blockade enhances expansion and function human tumor antigen-specific T cell responses in vitro and reduces tumor growth in cancer patients remains to be demonstrated. Although CD96 blockade together with other IR blockade are efficacious in multiple mouse tumor models, more data are needed to determine whether CD96 blockade alone or in combination significantly augment human tumor antigen-specific T cell responses in vitro. CD226 acts as a master regulator of dual PD-1/TIGIT blockade in cancer immunotherapy. Therefore, additional strategies are needed to prevent CD226 downregulation or augment its expression by TILs. To this end, combinatorial therapy with IL-15 and dual PD-1/TIGIT blockade may prove useful. Whether the antitumor effects of anti-TIGIT antibodies in cancer patients are Fc-dependent will need to be thoroughly investigated. Many ongoing clinical trials (Table 1) are testing Fc-engineered anti-TIGIT mAbs: IgG1 (MTIG7192/Genentech, MK-7684/Merck, and OP-313M32/Oncomed), inert-Fc IgG1 (BMS-986207/Bristol-Myers Squibb; AB-154/Arcus), and IgG4 (ASP8374/Potenza/Astellas).

Highlights.

• We discuss the role of TIM-3 as well as several members of the TIGIT pathway, including TIGIT, CD96 and CD112R as inhibitory receptors of both innate and adaptive immunity to cancer.

• We review the preclinical data as well as early-phase clinical trials supporting that targeting Tim-3, TIGIT, CD96 or CD112R together with PD-1 increase antitumor immunity and clinical responses.

• Tim-3, TIGIT, CD96 and CD112R represent attractive targets for cancer immunotherapy in combination with PD-1 blockade.

Abbreviations

AML

acute myeloid leukemia

ADAM

a disintegrin and metalloprotease

atezo

atezolizumab

BAT3

HLA-B associated transcript 3

CTLA-4

cytotoxic T lymphocyte-associated antigen 4

CEACAM1

carcinoembryonic antigen-related cell-adhesion molecule 1

DCs

dendritic cells

Gal-9

Galectin-9

HCC

hepatocellular carcinoma

HMGB1

high-mobility group box 1

HLH

lymphohistiocytosis

ICB

immune checkpoint blockade

IFN

interferon

Ig

immunoglobulin

ITIM

immunoreceptor tyrosine-based inhibitory motif

ITT

Ig tail-tyrosine

IL

interleukin

IR

inhibitory receptor

LAG-3

lymphocyte activation gene-3

LSC

leukemia stem cell

mAbs

monoclonal antibodies

MDSCs

myeloid-derived suppressive cells

NK

natural killer

NSCLC

non-small cell lung cancer

ORR

overall response rate

PBMCs

peripheral blood mononuclear cells

PD-1

programmed cell death 1

PD-L1

programmed death ligand 1

PtdSer

phosphatidylserine

SHIP

SH2-containing inositol phosphatase

SHP

Scr homology domain 2-containing tyrosine phosphatase

SNPs

single nucleotide polymorphisms

SPTCL

subcutaneous panniculitis-like T cell lymphoma

TCR

T-cell receptor

TIGIT

T cell immunoreceptor with immunoglobulin and ITIM domain

TIL

tumor-infiltrating lymphocyte

TIM-3

T-cell immunoglobulin and mucin domain-containing molecule-3

tira

tiragolumab

TME

tumor microenvironment

TNF-α

tumor necrosis factor-α

Tregs

regulatory T cells