Combinatorial blockade for cancer immunotherapy: targeting emerging immune checkpoint receptors

The differentiation, survival, and effector function of tumor-specific CD8+ cytotoxic T cells lie at the center of antitumor immunity. Due to the lack of proper costimulation and the abundant immunosuppressive mechanisms, tumor-specific T cells show a lack of persistence and exhausted and dysfunctional phenotypes. Multiple coinhibitory receptors, such as PD-1, CTLA-4, VISTA, TIGIT, TIM-3, and LAG-3, contribute to dysfunctional CTLs and failed antitumor immunity. These coinhibitory receptors are collectively called immune checkpoint receptors (ICRs). Immune checkpoint inhibitors (ICIs) targeting these ICRs have become the cornerstone for cancer immunotherapy as they have established new clinical paradigms for an expanding range of previously untreatable cancers. Given the nonredundant yet convergent molecular pathways mediated by various ICRs, combinatorial immunotherapies are being tested to bring synergistic benefits to patients. In this review, we summarize the mechanisms of several emerging ICRs, including VISTA, TIGIT, TIM-3, and LAG-3, and the preclinical and clinical data supporting combinatorial strategies to improve existing ICI therapies.


Introduction
The cancer-immunity cycle refers to the process wherein tumor antigen-reactive T cells undergo successful priming and differentiate into cytotoxic killer T cells that infiltrate tumor tissues and eliminate cancer cells (1).The differentiation, expansion, survival, and effector function of these tumor-specific cytotoxic T cells (CTLs) is regulated by the collective signaling effects of the T-cell receptor, costimulatory/ coinhibitory receptors, and cytokine receptors, which culminate in transcriptional and epigenetic programs to guide T-cell fate.Unlike in acute viral infections where effector CTLs and memory T-cell responses develop properly, tumor-specific CTLs exhibit dysfunctional states in response to chronic stimulation and a myriad of immunosuppressive factors in the tumor microenvironment (TME).These T cells progressively lose proliferative capacity, memory potential, and effector functions, and enter an "exhausted" state.Exhausted T cells upregulate the expression of multiple ICRs, including PD-1, CTLA-4, VISTA, TIGIT, TIM-3, and LAG-3, which sustain dysfunctional antitumor T-cell responses (2,3).
Immune checkpoint inhibitors (ICIs) are antibodies or small molecules that bind and block the function of ICRs, thereby reducing tumor-induced T-cell exhaustion and restoring anticancer immunity.Ipilimumab, the monoclonal antibody (mAb) blocking cytotoxic T lymphocyte antigen 4 (CTLA-4), was the first ICI therapy approved by the Food Drug Administration (FDA) in 2011.Currently, several mAbs targeting CTLA-4, PD-1, and PD-L1 have been approved for clinical applications.However, despite revolutionizing the field of oncology, the major challenge of existing ICI therapies is the overall low response rate.Understanding the unique molecular and cellular mechanisms of each ICR may support the development of novel combinatorial therapies that optimally restore antitumor immunity.
This review summarizes updated literature regarding the established and emerging ICRs: PD-1, CTLA-4, VISTA, TIGIT, TIM-3, and LAG-3.Due to the scope limitation, we omit discussions of additional emerging ICRs such as B7-H3, B7-H4, HHLA2, and butyrophilin-like 2 (BTNL2), which have been reviewed elsewhere (4).Herein, we provide an overview of each ICR's structure, expression, signaling mechanisms, and current preclinical and clinical data.We also elaborate on the concept that multiple ICRs operate concurrently to impair the expansion, survival, and effector functions of tumor-reactive cytotoxic T cells (Figure 1), as well as control the maturation and function of dendritic cells (DCs), macrophages, and myeloid-derived suppressor cells (MDSCs) (Figure 2).Given the frequent coexpression and functional crosstalk of these ICRs, we affirm the concept that combinatorial targeting of ICRs may achieve synergistic therapeutic outcomes compared to monotherapies.

Programmed death -1 PD-1 structure and expression
Programmed death -1 (PD-1, CD279) belongs to the B7/CD28 family of receptors, which are type-I transmembrane proteins consisting of an immunoglobulin variable (IgV) domain, a transmembrane domain, and a cytoplasmic tail with signaling capacities.PD-1 engages the ligands PD-L1 and PD-L2 and acts as a coinhibitory receptor that regulates both the adaptive and innate arms of the immune system (4,5).
PD-1 expression is detected in activated T cells, Foxp3 + regulatory T cells (Tregs), natural killer (NK) cells, innate lymphoid cells (ILC2s), B lymphocytes, macrophages, DCs, and monocytes.In T cells, PD-1 gene expression is induced by TCR signaling and positively regulated by multiple transcription factors including AP-1, NFATc1, FoxO1, NF-K B, Notch, STAT, and IRF9 (5).In cancers and chronic viral infections, PD-1 expression in exhausted T cells is significantly higher than in T cells from healthy hosts (3).The expression of PD-1 and its ligand PD-L1 on immune cells and cancer cells may serve as an indicator of disease progression and poor prognosis in a wide range of cancers (6).

Molecular mechanisms of PD-1
The intracellular domain of PD-1 contains an immunoreceptor tyrosine-based inhibitory motif (ITIM) and an immunoreceptor tyrosine-based switch motif (ITSM) (5).In T cells, the engagement of PD-1 by its ligand PD-L1 results in the recruitment of the tyrosine-protein phosphatases SHP1 and SHP2, which downregulate the phosphoinositide 3-kinase (PI3K), mitogenactivated protein kinase (MAPK), and mammalian target of rapamycin (mTOR) pathways.CD28 can be directly dephosphorylated by SHP2 and is the major target of PD-1 inhibitory signaling (7).At the cellular level, the consequences of the PD-1 pathway are multifaceted, resulting in altered T-cell metabolism with impaired glycolysis and augmented fatty acid oxidation, reduced cell expansion and effector cytokine production, and impaired T-cell mobility (3,4).
In addition to T cells, PD-1 is expressed in tumor-associated macrophages and inhibits their phagocytic function, which in turn controls antitumor immune responses (14).Furthermore, PD-1 plays a role in regulating tumor-driven emergency myelopoiesis.PD-1 deletion in myeloid progenitors reduced the accumulation of GMPs and MDSCs, which may be the result of elevated ERK1/2 and mTORC1 signaling and metabolic reprogramming (15).In preclinical models and cancer patients, blocking interactions of PD-1 with PD-L1 augments the effector function of PD-1 + exhausted CTLs, and induces the expansion of TCF1 + progenitorlike exhausted T cells with self-renewal capacity (16).On the other hand, blocking PD-1 may trigger hyperproliferation and suppressive function of Tregs and contribute to hyperprogressive diseases (17).

Targeting the PD-1/PD-L1 axis for cancer immunotherapy
Monoclonal antibodies specific for PD-1 (nivolumab, pembrolizumab), and PD-L1 (durvalumab, atezolizumab, and avelumab) have proven to be clinically effective and gained FDA approval across a wide range of cancers, such as skin cancer, lung cancer, Hodgkin lymphoma, renal cell carcinoma (RCC), head and neck cancer, bladder cancer, colorectal cancer, liver cancer, gastric cancer, triple negative breast cancer, and cervical cancer (18,19).Additional antibodies blocking PD-1, such as cemiplimab, camrelizumab, sintilimab, toripalimab, tislelizumab, zimberelimab, prolgolimab, and dostarlimab, have been approved for cancer applications worldwide.A meta-analysis of randomized controlled trials has concluded that anti-PD-1/PD-L1 inhibitors are more advantageous for treating advanced and metastatic cancers than conventional therapies, with better overall survival and progression-free survival particularly in male patients with younger age, without central nervous system or liver metastasis, no EGFR mutations, and with higher PD-L1 expression (18).
While PD-L1/PD-1 inhibitors are approved for treating an expanding list of cancers, their use as monotherapies generated an overall low response rate, due to mechanisms of primary and acquired resistance (20,21).To improve the response rate to ICIs, numerous combination strategies have been studied in preclinical and clinical trials, including combining PD-L1/PD-1 inhibitors with chemotherapeutics such as cyclophosphamide, radiotherapy, targeted therapy, agonistic costimulatory antibodies targeting CD134, CD137 or ICOS, innate immune stimulators such as STING agonists, epigenetic modulators, and cancer vaccines such as oncolytic viruses (19,22,23).On the other hand, these combinatorial regimens fail to address the roles of other nonoverlapping ICRs that constitute one of the dominant resistance mechanisms to PD-1/PD-L1 inhibitors.In the rest of this review, we will summarize studies of emerging ICRs (i.e., VISTA, TIGIT, TIM-3, and LAG-3) and demonstrate the rationales for combinatorial therapies targeting non-redundant ICRs together with PD-1/PD-L1 inhibitors.

Molecular mechanisms of CTLA-4
CTLA-4 inhibits the expansion, cytokine production, and differentiation of conventional T cells and contributes to the development and function of Foxp3 + Tregs.CTLA-4 exerts inhibitory effects by competing against CD28 due to its higher affinity for B7 molecules, as well as by recruiting phosphatases SHP2 and PP2A, which in turn downregulate signaling of TCR and CD28 (39-42).In addition to T-cell intrinsic mechanisms, CTLA-4 indirectly suppresses T-cell responses by modulating dendritic cells: CTLA-4 downregulates the surface expression of B7 molecules through trans-endocytosis (43) or induces the expression of indoleamine 2,3-dioxygenase (IDO), which in turn impairs T-cell proliferation (44).CTLA-4 also reverses the stop signal in activated T cells and reduces the contact time between T cells and APCs, leading to decreased cytokine production and T-cell proliferative responses (45).
The mechanisms of CTLA-4-mediated immunosuppression in cancers are distinct from PD-1 and potentially synergistic with PD-1 (46): although both receptors act on activated conventional T cells, PD-1 controls effector T-cell function at a later stage, mainly within peripheral tissue sites and the tumor microenvironment, while CTLA-4 intercepts T-cell priming in the lymph nodes and governs the function of Tregs (47,48).CTLA4 is constitutively expressed in Foxp3 + Tregs and CTLA-4-specific antagonistic antibodies not only augment effector T-cell activation but also induce ADCC-mediated depletion of tumor-infiltrating Tregs (49-51).On the other hand, unlike PD-1 and PD-L1, CTLA-4 is not expressed in myeloid cells and does not directly regulate suppressive myeloid cells within the TME.These functional distinctions provide mechanistic rationales for developing combination therapies targeting both axes.

Combinatorial blockade of PD-1 and CTLA-4
Studies have shown that while CTLA-4 and PD-1 blockade each boosts antitumor T-cell responses, dual blockade results in stronger therapeutic outcomes in preclinical models and human patients (52-54).ICI monotherapies induced the expansion of different tumor-infiltrating T cells (TILs), i.e., PD-1 blockade expanded exhausted-like CD8 + CTLs, whereas CTLA-4 blockade expanded both ICOS + Th1-like CD4 effectors and exhausted CD8 + CTLs.In contrast, the combined blockade induced the expansion of terminally differentiated effector CD8+ CTLs that are not seen in monotherapies and further increased Th1-like CD4 + effector T cells (52, 53).Similar findings have been shown in human melanoma patients treated with ipilimumab and nivolumab therapy.In addition to melanoma, dual blockade of CTLA-4 and PD-1 was studied in a murine breast cancer model (53).While monotherapies showed modest effects, combination therapy led to complete tumor regression in a majority of mice.The synergistic efficacy was due to the anti-CTLA-4 antibody-induced expansion of the T-cell receptor (TCR) repertoire and augmented functionality of TILs, accompanied by intratumoral Treg depletion.Taken together, these studies have demonstrated the mechanisms of synergy with dual ICI therapy that may guide clinical applications.
Ipilimumab (Yervoy) was the first FDA-approved monoclonal antibody for cancer immunotherapy, owing to robust clinical responses for metastatic melanoma (55, 56).We summarize recent clinical trials that have advanced PD-1 and CTLA-4 combinatorial therapy; comprehensive overviews of other clinical trials involving ipilimumab can be found in other reviews (57,58).
As a monotherapy, the effect of ipilimumab is not as strong as that of the PD-1 antibody nivolumab (Opdivo) for resected stage III or IV melanoma and showed shorter survival and higher toxicity for patients than the PD-1 antibody pembrolizumab (Keytruda) (59, 60).However, when ipilimumab was given concurrently with PD-1 antibody, dual blockade therapy demonstrated significantly improved outcomes in clinical studies.The advantages of dual ICI therapy were first noted in a Phase I dose-escalation study using nivolumab and ipilimumab administered together, which led to better response rates and progression-free survival compared to previously reported results from either monotherapy (61).A subsequent phase III study highlighted better responses and survival with combinatorial therapy when used for metastatic melanoma patients with PD-L1 negative tumors compared to either nivolumab alone or ipilimumab alone, despite the higher occurrence of grade 3 or 4 treatment-related adverse events (62).Follow-up studies showed durable responses and sustained benefits for survival in these patients across multiple years (63)(64)(65).Treatment-naive patients with advanced melanoma also benefited from nivolumab-plus-ipilimumab treatment, once again producing higher objective-response rates and progression-free survival with acceptable safety profiles compared to ipilimumab alone (57).
Current research continues to advance PD-1 and CTLA-4 combinatorial immunotherapy in the treatment of other cancers.Beyond melanoma, FDA approval of anti-PD-1 and anti-CTLA-4 dual therapy has expanded to hepatocellular carcinoma (HCC), unresectable pleural mesothelioma, RCC, metastatic non-small cell lung cancer (NSCLC), and advanced or metastatic esophageal squamous cell carcinoma (66)(67)(68).Combinatorial ICI therapy in the neoadjuvant setting has also shown promise, with tolerance and strong pathological responses for late-stage melanoma, early-stage colon cancers, and late-stage urothelial cancer (69,70).Dual blockade of CTLA-4 and PD-1 is currently being evaluated in numerous clinical trials for advanced solid tumors, such as head and neck squamous cell carcinoma (HNSCC) and glioblastomas (NCT04080804, NCT04606316).For testing combined treatment with pembrolizumab (anti-PD-L1), a randomized, double-blind phase III KEYNOTE-598 study (NCT03302234) showed that in patients with metastatic NSCLC, adding ipilimumab to pembrolizumab did not improve efficacy and exhibited greater toxicity than pembrolizumab monotherapy (71).Another phase I expansion trial (NCT02089685) evaluated the efficacy and safety of pembrolizumab combined with a reduced dose of ipilimumab in patients with advanced melanoma and RCC and showed manageable toxicity profile and robust antitumor activity (72).

VISTA VISTA structure and expression
V-domain immunoglobulin suppressor of T-cell activation (VISTA, alias Gi24, Dies-1, PD-1H, DD1a) is homologous to B7 family receptors and acts as a negative regulator of antitumor immunity and autoimmunity (73)(74)(75)(76)(77)(78).VISTA is a type I transmembrane protein containing a single IgV-like extracellular domain (ECD), a transmembrane segment, and a cytoplasmic tail that does not contain ITAM, ITIM, or ITSM motifs.Structural studies have revealed unique features of the VISTA ECD that are distinct from those of other Ig superfamily members, including two additional disulfide bonds, the insertion of an unstructured C-C' loop, the striking enrichment of histidine residues within the ECD, and an extra H b-strand that forms an intramolecular clamping disulfide bond (79, 80).Mutagenesis studies have demonstrated that these structural features contribute to the surface orientation and suppressive function of VISTA (79,80).
VISTA expression in mice is largely restricted within the hematopoietic compartment, with the highest expression on CD11b + myeloid lineages such as monocytes, macrophages, granulocytes, and dendritic cells (73,74).VISTA is also expressed in lymphocytes including NK cells, TCRgd T cells, naïve CD4 + and CD8 + TCRab T cells, and Foxp3 + Tregs.A similar expression pattern of VISTA is seen in human peripheral blood monocytic cells.VISTA gene expression is positively regulated by the transcription factors P53, HIF1-a, and STAT3 (81-83).However, whether VISTA exerts any impact on the functions of HIF1-a and STAT3 remains unknown.VISTA expression is also regulated by TGF-b/Smad3 signaling in T cells and myeloid cells (84).

Molecular mechanisms of VISTA
VISTA impairs antitumor immunity through its ligand activity in myeloid cells and T cell-intrinsic activity.Although it has been speculated that VISTA also acts as an inhibitory receptor (96), the signaling mechanism is unclear and it remains possible that T cell-intrinsic activity may rely on cis interactions with other signaling partners.At the molecular level, several partners, such as PSGL-1, VSIG3, and galectin-9, have been identified to engage VISTA (97)(98)(99).While PSGL-1 was suggested as an inhibitory receptor for VISTA, VSIG3 was considered a ligand.Galectin-9 binds VISTA and forms a protein complex that promotes galectin-9-mediated apoptotic signaling.At the cellular level, VISTA regulates the development and function of macrophages, MDSCs, neutrophils, TCRgd T cells, and CD4 + /CD8 + conventional T cells (74,75,78,100,101).In macrophages, VISTA impairs TLR signaling by regulating the ubiquitination and stability of TRAF6 (102).Blocking VISTA synergizes with a TLR-agonistic vaccine by augmenting the activation of DCs and macrophages, increasing the production of stimulatory cytokines such as IL-12 and IL-27, and promoting the effector function of tumor-specific CTLs.VISTA also contributes to the suppressive function of MDSCs, although the exact molecular mechanisms remain undefined (82,102).

Combinatorial blockade of VISTA and PD-1
In preclinical models, genetic deletion of VISTA or treatment with anti-VISTA mAb delayed tumor regression by inducing DC maturation, reducing the abundance of adaptive Foxp3 + Tregs, reducing the abundance of MDSCs, and augmenting the effector function and abundance of CTLs (73,76).
Studies led by Liu et al. first established the nonredundant and synergistic role of VISTA and PD-1 in mounting immune responses against self and tumor antigens (103).In both B16 melanoma and CT26 colon tumor models, combinatorial treatment with anti-VISTA and anti-PD-L1 mAbs resulted in tumor regression and long-term survival in comparison to monotherapies (103,104).A separate VISTA-blocking mAb, SG7, suppressed the interaction between VISTA and VSIG3 or PSGL-1 and showed efficacy in combination with PD-1 blockade in the MC38 colon tumor model (105).Finally, a unique role of VISTA in promoting naive T-cell quiescence has been identified (106).Accordingly, a study in a CT26 tumor model showed that a triple blockade of VISTA/PD-1/CTLA-4 could improve the efficacy of PD-1/CTLA-4 dual blockade by promoting antigen-presentation in myeloid cells and reducing the quiescent state of CTLs (107).
Several clinically relevant VISTA-blocking agents have been developed and entered clinical trials.VSTB112 (Janssen Inc) was the first anti-VISTA mAb tested in the clinic (NCT02671955).CA-170 (Curis Inc) is an orally available small molecule that has dual targeting activities against PD-L1/L2 and VISTA.In preclinical models, CA-170 rescued T-cell function similarly to PD-1 antagonists and inhibited the growth of B16 melanoma, CT26, and MC38 murine tumor models (108,109).CA-170 was tested in a phase I trial (NCT02812875) and a phase II trial (Clinical Trials Registry-India CTRI/2017/12/011026) (110).CA-170 showed an excellent safety profile and encouraging clinical activity in classic Hodgkin lymphoma and advanced NSCLC (109).HMBD-002 (Hummingbird Bioscience) is a human VISTA-specific mAb that binds to the C-C' loop of VISTA and blocks its interaction with VSIG3 (111).Studies of murine and humanized mouse models showed the effects of HMBD-002 in reducing MDSCs and augmenting T-cell responses.The phase I trial of HMBD-002 is ongoing (NCT05082610).W0180 (Pierre Fabre Inc) is a human VISTA-specific mAb being tested in a phase I trial (NCT04564417).The NCT05082610 and NCT04564417 trials will both test VISTA inhibitors in combination with pembrolizumab.KVA12123 (Kineta Inc) is a third human VISTA-targeting mAb that has recently been granted FDA acceptance for testing in phase I/II trials.

TIGIT TIGIT structure and expression
T-cell immunoreceptor with Ig and ITIM domains (TIGIT) is an ICR that contains an IgV-like ECD, a type I transmembrane domain, and a cytoplasmic domain with ITIM and ITT motifs (112).TIGIT is expressed on NK cells, CD4 + /CD8 + conventional T cells, and Foxp3 + Tregs.In T cells, TIGIT expression is upregulated following TCR activation and is sustained with increased exhaustion (112).
In human cancers, TIGIT gene expression was found to be upregulated in tumors and correlated with poor prognosis for KIRC, KIRP, LGG, and UVM cancers (113).TIGIT protein expression is abundant in CD4 + /CD8 + TILs and Tregs from a wide range of cancer types and is often correlated with poor clinical outcomes or resistance to ICI therapies (114).Coexpression of TIGIT and PD-1 on CD8+ TILs, which is associated with dysfunctional antitumor immune responses, has also been observed in cancers such as HCC, glioblastoma (GBM), acute myeloid leukemia, NSCLC, and melanoma (114).

Molecular mechanisms of TIGIT
TIGIT binds to three ligands CD112, CD113, and PVR (CD155), out of which CD155 exhibits the highest affinity (115,116).The TIGIT/CD155 interaction inhibits the functions of NK cells, T cells, and APCs.Phosphorylation of both the ITT and ITIM domains is required for the inhibitory signaling of TIGIT in NK cells and T cells, partly by recruiting the adaptors Grb2 and SHIP1, which in turn dampen the PI3K, MAPK, and NF-K B signaling pathways (117,118).TIGIT also outcompetes CD226 for binding to CD155 and disrupts the costimulatory signal from CD226 in T cells (119).In addition to effector T cells, TIGIT is expressed in Foxp3 + Tregs and plays a role in promoting their differentiation, stability, and suppressive function (120-122).
In APCs such as DCs and macrophages, CD155 is phosphorylated upon engaging TIGIT and subsequently inhibits MAPK signaling, resulting in tolerogenic APCs that produce elevated levels of IL-10 but reduced levels of IL-12, and fail to properly stimulate cognate T cells (123).Another recent study demonstrated that leukemia-associated macrophages express TIGIT and that blocking TIGIT drives M1-like phenotypes and increases phagocytosis (124).

Combinatorial blockade of TIGIT and PD-1
The efficacy of the dual blockade of TIGIT and PD-L1 has been demonstrated in murine breast and colon carcinoma models (112).The combination therapy rejuvenated tumor-specific CD8 + CTLs by augmenting their expansion, effector functions, and the development of memory responses (112).A recent study has shown that the PD-1 and TIGIT pathways converge to regulate CD226, as both receptors impair the phosphorylation of CD226 (125).Furthermore, when CD8 + TILs from human liver cancers were treated with TIGIT and PD-1-blocking mAbs, the coblockade of TIGIT and PD1 significantly improved the expansion, cytokine production, and cytotoxicity of CD8 + TILs compared with single PD-1 blockade (126).Similar results were seen in an adoptive T-cell transfer study to treat human lung cancer, where dual blockade of TIGIT/PD-1 or TIM-3/PD-1 resulted in greater tumor control than PD-1 monotherapy (127).Together, these studies provide a strong rationale for blocking both the PD-1 and TIGIT pathways to allow optimal CD226-dependent costimulatory signaling in CD8 + T cells.

TIM-3 TIM-3 structure and expression
T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), along with TIM1 and TIM4, belongs to the TIM family of immunoregulatory proteins.The ECD of TIM-3 contains an immunoglobulin variable domain that binds to several ligands: galectin 9, phosphatidylserine, CEACAM1, and HMGB1 (136).Following the ECD is a mucin domain, a transmembrane domain, and a cytoplasmic domain that does not contain canonical inhibitory signaling motifs such as ITIM or ITSM motifs.
TIM-3 is also expressed in DCs, where its ligation induces the activation of Bruton's tyrosine kinase and c-Src, which inhibit NF-kB activation and subsequently reduce DC activation (158).In macrophages, TIM-3 has been reported to promote M2-like polarization by inducing SOCS1 (159).In monocytes and DCs, TIM-3 inhibits the cellular responses to TLR signaling and reduces the production of proinflammatory mediators (160).In a breast cancer model, blocking TIM-3 augmented the production of a key chemokine CXCL9 from CD103 + DCs, thereby improving the a

Combinatorial targeting of TIM-3 and PD-1
In preclinical models, dual blockade of TIM-3 and PD-1 restored the function of both CD4 + and CD8 + T cells and led to complete tumor regression whereas either monotherapy was not effective (162,163).A recent study has shown that PD-1 binds galectin-9 and that PD-1/TIM-3/galectin-9 may crosslink and form a lattice.As such, PD-1 functions to attenuate galectin-9/TIM-3induced apoptosis (164).It should be noted that VISTA also binds to galectin-9 and augments the inhibitory effects of TIM-3 (99).Thus, these findings may provide a rationale for future studies to test the combined blockade of PD-1, TIM-3, and VISTA, to improve the persistence and functions of tumor-reactive PD-1 + TIM-3 + CTLs.
In human cancers, TIM-3 and PD1 are often coexpressed on CD8 + T cells and mark the most dysfunctional T cell subsets.An earlier study of advanced melanoma showed that NY-ESO-1specific PD-1 + CD8 + TILs upregulate TIM-3 expression, which is correlated with dysfunctional phenotypes (165).Blocking TIM-3 augmented cytokine production and proliferation of T cells, while combined blockade of both TIM-3 and PD-1 showed synergistic effects.Similar findings were reported in colorectal cancer, where TIM-3 + PD-1 + CD8 + TILs represented the predominant fraction of TILs and targeting both TIM-3 and PD-1 enhanced cell expansion, cytokine production, and cytotoxic activity (166).Recent studies of diffuse large B-cell lymphoma found that TIM-3 + PD1 + TILs exhibited a transcriptomic signature of T-cell exhaustion, reduced proliferation, and impaired cytokine production, but these dysfunctions were restored by the blockade of PD1 or TIM-3 (167,168).Although there have not been any FDA-approved therapeutics targeting TIM-3, the pipelines for novel TIM-3 inhibitors are expanding: several TIM-3-specific antibodies (i.e., cobolimab, MBG453, Sym-023, BMS-986258, AZD7789, INCAGN02390, etc.) or TIM-3/PD-1 bispecific antibodies are being tested in clinical trials (169).A phase I/II trial (NCT02608268) evaluated MGB453 (anti-TIM3) in combination with PDR001 (anti-PD-1) in advanced solid cancers such as melanoma and NSCLC and showed excellent safety profile and preliminary antitumor activity (170).Similar encouraging results were shown by trials (NCT02817633 and NCT03680508) that evaluated TSR-022 (anti-TIM3) in combination with PD-1 inhibitors (171,172).In addition, a phase Ia/b trial evaluated the safety, pharmacokinetics, and efficacy of LY3321367 (anti-TIM3) plus LY3300054 (Anti-PD-L1) and showed modest antitumor activity (173).

LAG-3 LAG-3 structure and expression
Lymphocyte activation gene 3 (LAG-3, CD223) is an Ig superfamily ICR and is homologous to CD4 (174,175).The ECD of LAG-3 contains four IgV or IgC-like domains that are involved in ligand binding.The cytoplasmic domain of LAG-3 contains a serine phosphorylation site, the conserved KIEELE motif, and the glutamate-proline dipeptide repeat motif that is involved in its inhibitory signaling (176).

Combinatorial targeting of LAG-3 and PD-1
Preclinical studies have established that LAG-3 cooperates with PD-1 in controlling antitumor immunity (175,209).The striking synergy between LAG-3 and PD-1 has been demonstrated in murine melanoma, colon cancer, and ovarian tumor models, where the dual blockade against LAG-3 and PD-1 effectively controlled tumor progression that was resistant to respective monotherapies (205, 210).A study in the MC38 colon cancer model has shown that PD-L1 blockade elevated the expression of both costimulatory receptors (ICOS) and coinhibitory receptors (LAG3 and PD-1) in TILs, thereby providing a new mechanistic rationale for coblocking LAG3 (211).

Conclusions
Since the first FDA approval of ICIs in 2011, significant progress has been made toward optimizing existing ICI therapies.Taking the lessons from existing ICIs that target PD-1, PD-L1, and CTLA-4, current efforts in the field focus on identifying and targeting nonredundant ICRs that may potentially synergize with existing therapies.VISTA, TIGIT, TIM-3, and LAG-3 represent such candidates in the pipeline.Recent advances in understanding the converging role of ICRs in driving the dysfunction of both APCs and T cells (Figures 1, 2) have set the conceptual foundation for developing combinatorial therapies targeting these ICRs.Based on the frequent coexpression of ICRs in tumor tissues and the distinct yet convergent mechanisms of action (Table 1), it is expected that combined blockade of these emerging ICRs with PD-L1/PD-1 will result in additive or synergistic outcomes.Indeed, many novel ICI combination therapies are being investigated in early-stage trials (Table 2).To advance this concept into clinical applications, the field still faces some challenges, such as defining the molecular pathways and hierarchy of emerging ICRs, identifying the optimal ICR combinations for distinct cancer types and discrete biomarkers, and developing better preclinical models that present the full extent of immune-related toxicities as seen in human patients.In

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers.Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

FIGURE 1
FIGURE 1 Overview of coinhibitory ICRs and their effects in conventional T cells.T-cell activation requires TCR recognition of cognate antigens presented on APCs and costimulation provided by B7/CD28 or CD115/CD226 interactions.On the other hand, many coinhibitory ligand/receptor pathways are activated to dampen T-cell responses.The B7/CTLA-4 and PD-L1/2/PD-1 pathways are the cornerstones of the immune checkpoint paradigm.Emerging inhibitory ICRs, including TIGIT, LAG-3, TIM-3, and VISTA, each recognized by multiple ligands, play nonredundant yet convergent roles as the "brakes" of T-cell responses.

FIGURE 2
FIGURE 2 The signaling effects of ICRs in antigen-presenting cells.Aside from suppressing T-cell activation, many ICRs regulate the maturation, antigen presentation, cytokine production, and other effector functions of DCs and tumor-associated macrophages.CTLA-4 reduces the surface expression of B7 molecules through trans-endocytosis.LAG-3 and TIGIT trigger signaling in a reverse direction by engaging their respective binding partners MHCII and CD155.On the other hand, PD-1, TIM-3, and VISTA are expressed in APCs and transmit inhibitory signals to inhibit the effector functions of APCs, including phagocytosis, antigen presentation, and cytokine production.Both PD-1 and VISTA are also expressed in tumor-driven MDSCs and contribute to the differentiation and suppressive function of MDSCs.

TABLE 1
Blocking individual ICRs augments antitumor immune responses by convergent cellular and molecular mechanisms.

TABLE 2
Clinical trials testing combined targeting of ICRs., we emphasize that antitumor immunity is controlled by multiple nonredundant ICRs that together maintain immune dysfunction.Recent preclinical and early clinical data strongly support the rational design of novel ICI combinations to achieve synergistic therapeutic efficacies with manageable toxicities. conclusion