2.1 Structure, expression, and ligands

TIGIT is a transmembrane glycoprotein consisting of an extracellular Ig variable region, a type 1 transmembrane region, a cytoplasmic tail containing an immunoreceptor tyrosine-based inhibitory motif (ITIM), and an Ig tail tyrosine (ITT)-like motif.^8-13 It is highly expressed on various types of immune cells (Table 1).⁸ Numerous studies have proven that TIGIT and several other molecules act together to form an immune regulatory axis and lead to stimulatory or inhibitory signals, despite some mechanisms of the pathway that has been not fully understood yet. To date, the exact involvement of CD155, CD112, Nectin-4, CD226, TIGIT, CD96, CD112R, CD113, CD111, and the bacterial ligand Fap2 in this regulatory axis has been identified (Figure 1), which we will describe in detail below.

TIGIT binds to CD155 and CD112 to deliver inhibitory signals, with a higher affinity to the former.^14-17 CD155 is lowly expressed in normal human cells such as dendritic cells (DCs), endothelial cells, but highly in tumors such as lung cancer, melanoma, and so on. In 2020, Adi Reches et al. first reported that Nectin-4 could bind specifically to TIGIT and inhibit the activity of NK cells.¹⁸ Besides, CD113 is also a ligand for TIGIT but their interactions are unclear⁸ currently.

CD226, which expresses on T cells, monocytes, and NK cells,¹⁹ can compete with TIGIT and bind to CD155 and CD112 to transmit the costimulatory signals^{11, 13, 20} by enhancing the cytotoxicity of NK cells²¹ and synergizing with lymphocyte function-associated antigen-1 (LFA-1) to regulate T-cell functions.²² Besides, CD96 (also known as TACTILE) can also bind to CD155 with a higher affinity than CD226 but weaker than TIGIT.^{8, 23} Compared to TIGIT, CD96 possesses a more complicated extracellular structure including three Ig-like structural domains.²⁴ A YXXM motif of CD96 indicated a potential ability for activation,²⁴ which is confirmed by several studies.^{25, 26} Furthermore, CD112R was also a novel immune checkpoint on human T cells²⁷ and NK cells,²⁸ it binds to CD112 and plays a nonredundant synergistic role with TIGIT in immunosuppression.²⁹

Combining the existing studies, we conclude that TIGIT, CD96, CD112R, Nectin-4, and CD226 could bind to the same ligands but opposite functions.³⁰ Compared with the well-defined classical CD28/CTLA4-CD80/CD86 pathway³¹ which only regulates T-cell activation, CD226/CD96/TIGIT/CD112R/Nectin-4/CD155/CD112 pathway can affect not only T cell but also NK cell functions,³⁰ suggesting the regulatory ability in both innate and acquired immunity.

2.2 Mechanisms of action

Analysis of the crystal structure revealed that the TIGIT/CD155 interface displayed a “lock and key” interaction. A heterotetrameric structure consisting of a TIGIT homodimer and a CD155 homodimer is essential for inhibitory signal generation (Figure 2).³² Both TIGIT and CD155 contain an ITIM in the cytoplasmic tail,⁸ suggesting the ability to transmit inhibitory signals in the corresponding cells. The current uncovered mechanisms of TIGIT/CD155-mediated coinhibitory signaling included: direct intracellular signaling, reduction of costimulatory signaling, and cell-extrinsic signaling cascades (Figure 3).

3.1 NK cells

The expression of TIGIT was upregulated on NK cells in animal models as well as various human malignancies and may be associated with a poor prognosis.^{34, 50} In two tumor-bearing mouse models, TIGIT was associated with NK cell exhaustion characterized by less production of IFN-γ, and blocking TIGIT prevented the depletion and enhanced innate immunity through an NK cell-dependent manner.^{5, 50} Besides, in patients with multiple cancers such as chronic myelogenous leukemia, myelodysplastic syndromes, soft tissue sarcoma, and osteosarcoma, TIGIT expression was significantly upregulated in peripheral blood NK cells^51-53 and associated with a worse prognosis, suggesting that targeting TIGIT is an effective therapeutic strategy to enhance innate antitumor immunity and complement existing treatments.

Notably, several research groups have found that CD96, which also binds to CD155, acted as a suppressor in both mouse and human NK cells. However, limited studies also observed that CD96 could enhance NK cell functions in human malignancies,^{25, 26, 54} possibly owing to the coexistence of activating and inhibitory sequences of CD96 in humans.

3.2 Tumor-infiltrated CD8⁺ T cells

High expression of TIGIT was also seen on CD8⁺ TILs in both animal models and patients with solid tumors such as nonsmall cell lung cancer (NSCLC), melanoma, papillary and anaplastic thyroid cancer.^{13, 53, 55-60} By reducing proliferation and production of effective cytokines, the antitumor response was diminished. In some nonsolid tumors such as lymphoblastic leukemia and follicular lymphoma patients,^{37, 61} results as previously described were also observed. High endogenous expression of TIGIT on CD8⁺ TILs reduced secretion of IFN-γ and IL-17 and suggested that TIGIT blockade could restore the effector function of CD8⁺ T cells. This was also confirmed by the fact that in acute myeloid leukemia patients, siRNA knockdown of TIGIT could result in a significant increase in TNF-α and INF-γ and reduced sensitivity in CD8⁺ T cells to apoptosis.⁶²

Additionally, several researchers have found a decreased expression of CD226 along with the expression of TIGIT in CD8⁺ T cells,^{55, 62} thus a reduction of the CD226⁻mediated costimulatory signals. Also, TIGIT expression in Tregs and APCs could alter the TME, thereby indirectly inhibiting the antitumor response of CD8⁺ TILs. Similar to NK cells, CD8⁺ TILs were recently found accessible to both costimulatory and coinhibitory signals from CD96,^{63, 64} which needs further research to explain the mechanism.

3.3 Treg cells

A high proportion of Tregs was found in the TME in different types of tumors, which was characterized by highly upregulated TIGIT expression.^45-48 Compared with TIGIT⁻ Tregs, TIGIT⁺ Tregs upregulated several Treg-associated markers which are essential to the differentiation of naive T cells to Tregs phenotype. Several studies have verified that TIGIT⁺ Tregs selectively inhibit Th1/Th17 rather than Th2 immune responses via FGL2,^{12, 65, 66} leading to polarization of type 2 tumor-associated macrophages (TAM2),¹² which deprived nutrients in the TME and suppressed T cell functions. Meanwhile, TIGIT⁺ Tregs can also reduce the ability of CD155 to bind CD226 on CD8⁺ Teffs and NK cells,⁶⁷ thus affecting the Teffs and NK cell-mediated antitumor responses. Notably, it was recently found that a higher ratio of TIGIT/CD226 on Tregs was significantly associated with poor clinical outcomes toward ICB therapy.⁶⁸ Further researches should be carried out to illuminate the potential value of TIGIT/CD226 ratios in predicting ICB therapy clinical response.¹⁷

3.4 DCs

The binding of TIGIT to DCs can induce phosphorylation of the inhibitory ITIM motif of CD155 on DCs, triggering a signaling cascade that promotes a tolerogenic phenotype of DCs characterized by increased interleukin (IL)-10 secretion.⁸ As an immunosuppressive cytokine, IL-10 directly acts on T cells to suppress their proliferation and production of pro-inflammatory cytokines such as IFN-γ.⁸ Besides, DCs were also regulated by TIGIT⁺ Tregs to have a tolerogenic phenotype. This mechanism plays a synergistic effect with direct TIGIT/CD155 signaling in DCs to form a positive feedback loop amplifying a suppressive DC phenotype, inhibiting the generation of effector T cell responses.⁶⁷

3.5 Bacterium-dependent mechanism of immune escape in tumor

Unexpectedly, TIGIT has been reported to be involved in a bacterium-dependent tumor-immune evasion mechanism.⁶⁹ Fusobacterium nucleatum, an intrinsic oral bacterium, is a common anaerobic Gram-negative bacterium in the TME and negatively correlates with T cell density in colorectal cancer tissues, suggesting a possible role of this bacterium in tumorigenesis.⁷⁰ Gur et al. found that the functions of TIGIT⁻expressing NK cells and T cells could be inhibited when the Fap2 protein of F. nucleatum is bound to TIGIT, revealing a new bacterium-dependent mechanism of immunosuppression.⁶⁹

To sum up, TIGIT may involve multiple immune components of the antitumor immune responses, and its expression on various types of immune cells determines the involvement in innate and adaptive immunity. It suggested targeting TIGIT is a promising therapeutic strategy for cancers.

4.1 Broader specturum of cell-regulation

Kurtulus et al. found that TIGIT deficiency in Tregs was sufficient to retard tumor growth and could increase IFN-γ, TNF-α, and IL-2 productions by CD8⁺ TILs in an experimental model,⁴⁵ indicating that TIGIT expression in Tregs may perform a more prominent function in suppressing antitumor immune responses compared with it in CD8⁺ T cells. One possible explanation is that both CD96 and TIGIT are expressed on CD8⁺ Teffs, and the absence of TIGIT on the cells may also be replaced by CD96, whereas in Tregs, the lack of CD96 expression leads to a more nonredundant regulatory role of TIGIT.⁶⁷ Someone suggested that Tregs are of utmost significance in the initial phases of tumor development, as they initiate an immunosuppressive landscape during tumor development.⁷³ Some researchers have found that Tregs promoted dysfunctional phenotype of CD8⁺ TILs by modifying the TME.^{74, 75} However, TIGIT expression on CD8⁺ T cells may have a more pronounced effect in the established tumors. Kurtulus et al.⁴⁵ found TIGIT⁺ CD8⁺ TILs were not only poor mediators of tumor clearance, but also contributed to immune suppression locally in the TME.

Meanwhile, two recent studies have provided evidence for TIGIT expression in NK cells-mediated tumor metastasis.^{76, 77} Besides, recently additionally several studies on TIGIT in B cells also suggest its possible involvement in some B cell-mediated immune processes.^78-80 As elaborated, in contrast to the PD-1 pathway, the involvement of TIGIT in NK cells and B cells makes it possible to regulate innate immunity, adaptive immunity, and humoral immune, which functions as a nonredundant complement to PD-1 inhibition.

4.2 Higher tissue specificity and safety

CTLA-4 and PD-1 stand for first line of immune checkpoints and are primarily responsible for maintaining self-tolerance and restriction of T cells, the absence of which may lead to severe autoimmune reactions, whereas the deficiency of TIGIT and several other molecules does not, as their suppressive capabilities are just apparent in a susceptible background or actively induced diseases. Notably, compared to others, a unique regulatory role of TIGIT is to alter the balance between type 1/17 and type 2 immunity.⁸¹ Furthermore, TIGIT seems to be expressed more frequently in TME cells than that in peripheral, theoretically providing well-targeted treatment to reduce systemic autoimmune toxicity.⁸² For example, TIGIT^−/−35 or CD96^−/−83 mice have not shown significant spontaneous development of immunopathology, which is distinct from ICB targeting CTLA-4 and PD-1. Nevertheless, it has also been shown that TIGIT^−/− mice are more sensitive to the induction of experimental autoimmune encephalomyelitis (EAE)³⁵ and TIGIT blockade could increase the development of experimental arthritis,⁹ suggesting the results of animal experiments need to be interpreted cautiously.

4.3 Coblockade of TIGIT and other immune checkpoints and clinic trials

TILs usually upregulate TIGIT expression together with PD-1, TIM-3, and LAG-3,⁸⁴ and coblockade of TIGIT and other checkpoints such as PD-1,⁵⁵ PD-L1,^{13, 85} and TIM-3⁴⁵ have been shown to specifically enhance T cell function. While every single blockade did not significantly hinder colorectal tumor growth in mice, TIGIT and PD-1 inhibition together synergistically enhanced the proliferation and the antitumor function of CD8⁺ TILs and prolonged overall survival.^{13, 50, 86}

Several recent studies have shed light on the mechanism of dual blockade of the PD-1 and TIGIT coinhibitory receptors in antitumor therapy. In preclinical tumor models, Banta et al. found that PD-1 inhibited phosphorylation of both CD226 and CD28 on CD8⁺ T cells via its intracellular effectors such as SHP-2 phosphatase, which was recruited by the PD-1 ITIM-containing intracellular domain, whereas TIGIT restricting CD226 costimulation by blocking interaction with CD155.⁸⁷ Therefore, coblockade of TIGIT and PD-1 is required to completely restore CD226 signaling, resulting in optimal antitumor CD8⁺ T-cell responses.⁸⁷ Other researchers have established hepatocellular carcinoma (HCC) mouse models and found that the use of PD1 mAb simultaneously increased the expression of immune checkpoint molecules, including TIGIT, which may thereby lead to a depleted state of CD8⁺ T cells and ultimately to resistance to PD1 mAb. Moreover, high levels of TIGIT expression in peripheral blood T cells of HCC patients unresponsive to PD1 mAb have also been observed. Targeted blockade of TIGIT may sensitize HCC to PD1 mAb, considering that TIGIT is associated with PD1 mAb resistance⁸⁸ and coblockade an optimal therapy. Recently, a preclinical study found that CD155/TIGIT axis played an important role in mediating immune evasion of pancreatic cancer, and coblockade of TIGIT/PD-1 coupled with CD40 agonist produced a strong antitumor response.⁸⁹ Based on the results of some animal models, clinical trials on the combination of PD-1/PD-L1 and TIGIT blockade are being widely conducted in recent years. Moreover, recently a phase I trial about a recombinant humanized TIGIT×PD-L1 bispecific antibody in advanced solid tumors was started (NCT05102214) and we are looking forward to its findings.

Additionally, a prerequisite for PD-1/TIGIT coblockade for tumor therapy may be the high expression of CD226 on lymphocytes. Previous studies found that PD-1/TIGIT coblockade was abolished when CD226 was blocked, indicating that ICB targeting TIGIT may function mainly via CD226 costimulatory signals.^{13, 55, 90} Jin et al.⁹¹ verified that CD226^hi CD8⁺ T cells are required in a blockade of TIGIT pathway. Moreover, they also found that CD226⁺ CD8⁺ T cells in patients with pancreatic carcinoma might potentiate the effects of TIGIT or PD-1 blockade.⁹¹ Nevertheless, CD8⁺ TILs tend to decrease the expression of CD226 in tumors, partly limiting the effectiveness of coblockade of PD-1 and TIGIT.^{55, 62} So we expected the combination therapy (TIGIT blocker + CD226 agoist) in clinical trials.

Given the emerging advantages of TIGIT and ample preclinical evidence, many pharmaceutical agencies have set up clinical trials of anti-TIGIT mAbs as single agents or in combination in multiple types of carcinomas in recent years, the current (by 2022) drugs entering clinical trials include 18 drugs, with a total of 109 programs, and 17 of them have entered Phase III clinical trials in various solid tumors (Table 2). Here, we summarized the related clinical trials.

5.1 Autoimmune disease and transplant status

The central role of coinhibitory receptors in maintaining immune system homeostasis makes it possible that their loss may lead to spontaneous, severe autoimmunity (Figure 6). These days, increasing studies have illustrated the function immune checkpoint played in autoimmune diseases, and its upregulation may contribute to the prevention and management of autoimmune diseases.

5.2 Chronic infection

Persistent antigen exposure and dysfunctional Teffs are essential features of chronic infection.⁴ When the infection antigens persist, the immune system will initiate a number of suppressive pathways to inhibit the response and prevent tissue damage from immune-mediated, a process known as “depletion.”¹⁰⁴ In fact, T cells express various immune checkpoints in chronic infections, including CTLA-4, PD-1, TIGIT, TIM-3, and LAG3.^105-109 Limited therapies and a significant population of patients have triggered studies of the coinhibitory receptors in chronic infection diseases^110-115 such as human immunodeficiency virus (HIV), hepatitis B virus (HBV) and so on.¹¹⁶

5.3 Human pregnancy and pregnancy-related diseases

Previous studies have clarified that Tregs were vital in embryo implantation, placenta formation, and maternal-fetal tolerance induction.¹²⁴ Recently, a research reported that TIGIT⁺ Tregs had a unique ability to establish immune tolerance at the human maternal-fetal interface.¹²⁵ Another study also showed that the TIGIT gene was upregulated in uterine Tregs (uTregs), especially at the placental bed site.¹²⁴ Wang et al. found that TIGIT expression on peripheral blood NK (pNK) cells increased gradually during pregnancy, along with a decreased expression of granzyme B, IFN-γ as well as CD107a in pNK cells.¹²⁵

In fact, several studies have verified that TIGIT low expression is related to adverse pregnancy outcomes and may be a potential target for therapy. A study about unexplained recurrent miscarriage patients and normal pregnancy showed that an increased expression of TIGIT in the decidual γδT cells significantly plays essential roles in maintaining pregnancy,^{126, 127} which was also confirmed in a study of patients with pre-eclampsia.¹²⁸ Fu et al. detected the expression of TIGIT in decidual immune cells at the maternal-fetal interface. Then by generating recombinant TIGIT-Fc fusion proteins, they confirmed that treatment with TIGIT-Fc of human decidual immune cells could increase IL-10 production, leading the decidual CD4⁺T cells to a Th2 phenotype, suggesting that TIGIT-Fc fusion protein has therapeutic potential in pregnancy failure.¹²⁹ All the existing studies indicated that treatment targeting TIGIT pathway has seen promising clinical benefits in adverse pregnancy outcomes, further studies are needed to better understand its function in maintaining maternal-fetal tolerance and the potential value in the diagnosis and treatments of pregnancy-related diseases (Figure 7).