LAG-3 is expressed on T cells and interacts with a variety of ligands to inhibit T cell activity, as shown in Figure 1 . MHC-II is a natural ligand for LAG-3, and their binding has an essential effect on the behavior of T cells and tumor cells (42). On the one hand, the binding of LAG-3 to MHC-II affects anti-tumor immune responses by negatively regulating the activity of T cells (42, 43). In detail, LAG-3 binds to MHC-II molecules with high affinity and blocks T cell receptor (TCR) signaling, thereby inhibiting T cell proliferation, activation, and cytokine secretion (32, 33) ( Figure 2 ). This inhibitory effect not only affects CD4⁺ and CD8⁺ T cells but also promotes the expansion of Tregs in the TME, which have a stronger immunosuppressive function and can further limit anti-tumor immune responses through cell contact (35, 44).

On the other hand, interaction of LAG-3 with MHC-II activates several survival pathways within tumor cells, including MAPK/Erk and PI3K/Akt signaling pathways (45, 46). Activating these pathways enhances tumor cell resistance to apoptosis, especially in MHC-II-expressing melanoma cells (46). It has been shown that LAG-3 binding to MHC-II protects tumor cells from Fas-mediated apoptosis and drug-induced apoptosis, such as etoposide, by activating the MAPK/Erk pathway and the PI3K/Akt pathway (47).

Overall, the binding of LAG-3 to MHC-II affects immune escape mechanisms in two ways. On one hand, it inhibits the immune surveillance function of T cells, and on the other hand, it enhances the survival of tumor cells. This interaction makes tumor cells more resistant in the face of the host immune system and therapeutic interventions, suggesting that strategies targeting LAG-3 and MHC-II interaction may have substantial clinical applications in cancer therapy.

Galectin-3 (Gal-3), or lectin-3, is a glycoprotein widely expressed in immune cells, including macrophages, dendritic cells, and NK cells (30) ( Figure 1 ). Studies have shown that Gal-3 is able to induce T cell anergy through TCR aggregation, thereby inhibiting T cell function (48). This inhibition has been observed in a variety of cancers, such as endometrial cancer, vulvar squamous cell carcinoma, and multiple myeloma (49–51). Furthermore, the function of these inhibited T cells could be restored by the knockout of surface Gal-3. Kouo et al. found that LAG-3 expression in the TME is required for Gal-3 to exert its inhibitory function (52). In contrast to LAG-3⁺ CD8⁺ T cells, LAG-3 KO CD8⁺ T cells were able to secrete more IFN-γ in the presence of Gal-3 (52). These results suggest that the interaction of LAG-3 with Gal-3 is an essential mechanism for regulating CD8⁺ T cell function.

LSECtin is a C-type lectin-like receptor typically expressed in the liver and has inhibitory immunomodulatory effects (53) (Figure 1). Xu et al. found that LSECtin can be expressed on the surface of murine B16 melanoma cells, inhibiting tumor-specific T cell immune responses and promoting the growth of tumor cells (53). Blockade of LAG-3 significantly improved LSECtin-mediated IFN-γ secretion by CD8⁺ T cells (53). However, the study by Xu et al. only preliminarily demonstrated the effect of LSECtin and LAG-3 interaction on tumor cells. The specific mechanism and other functions still need to be further explored.

FGL-1 is another ligand for LAG-3, expressed in hepatocytes, tumor cells, and some immune cells (54) (Figure 1). FGL-1 secreted by tumor cells within the TME can bind to LAG-3 on TILs, thereby inhibiting the anti-tumor activity (55). An in vivo experiment by Wang et al. showed that anti-FGL-1 mAb effectively activated T cell function and enhanced anti-tumor effects in WT mice. However, in LAG-3-deficient mice, the addition of FGL-1 mAb had no effect on the activation of T cells (56). They suggested that the interaction between LAG-3 and FGL-1 was a key mechanism leading to T cell inhibition. When anti-FGL-1 antibody was added, the inhibitory signaling pathway was blocked and T cell function was restored in WT mice. However, the LAG-3 KO mice model lost its ability to suppress T cells, so that T cell activation did not change significantly after adding anti-FGL-1 antibody. Furthermore, FGL-1 showed a more significant inhibitory effect on 3A9 T cell lines with IL-2-induced LAG-3 overexpression (56). The addition of anti-FGL-1 mAb to 3A9 T cell lines resulted in positive regulation of TNF-α and an increase in IFN-γ levels, which restored T cell activation (56). These results suggest the importance of the interaction between LAG-3 and FGL-1 in T cell inhibition and tumor development.

Recently, Guy et al. have identified the TCR-CD3 complex as a potential novel ligand for LAG-3 (57) (Figure 1). Immuno-tyrosine-based activation motifs (ITAMs) in the CD3 complex are phosphorylated when the TCR identifies and attaches to the antigenic peptide-MHC complex (58). This phosphorylation, mediated by Lck (Src family tyrosine kinase), activates downstream signaling molecules such as ZAP70, LAT, and Syk, ultimately resulting in T cell activation, proliferation, and differentiation (59). This complex forms a tight contact area called the immune synapse. Guy et al. indicated that LAG-3 can bind to TCR-CD3 complexes and then migrate to immune synapse in CD4⁺ and CD8⁺ T cells in the absence of MHC-II (57). They suggested that conserved acidic tandem glutamic acid-proline repeats in the cytoplasmic tail of LAG-3 decrease the PH of the immune synapse, resulting in the dissociation of the tyrosine kinase Lck from CD4 or CD8 coreceptors and preventing TCR signaling and T-cell activation (57, 60). Workman et al. pointed out that the highly conserved ‘KIEELE’ motif in the cytoplasmic domain of LAG-3 is essential for its function (61). However, there is still no breakthrough in the study of this motif.

In summary, as an emerging immune checkpoint, the study of its interaction with ligands is still at an initial stage. Excluding MHC-II, the binding mechanism of other ligands to LAG-3 and the function of tumor immunity still need to be elucidated. Targeting LAG-3 and its ligands is one of the effective therapies to enhance anti-tumor immune responses and inhibit tumor growth. In-depth study of its interaction with ligands is essential to understand and elucidate the LAG-3-mediated immune inhibition.

Several studies have shown that LAG-3 plays an essential role in immunosuppression and tumor growth, indicating that inhibiting LAG-3 could have positive therapeutic effects in cancer treatment (47). For example, in vitro studies have shown that inhibition of LAG-3 signaling in TILs from patients with melanoma restores the ability of CD4⁺ and CD8⁺ T cells to secrete IFN-γ, thereby enhancing their antitumor activity (25, 26). Furthermore, blocking LAG-3 on circulating NK cells from lung cancer patients increased cell cytotoxicity and IFN-γ levels (72). Maruhashi et al. suggested that the MHC-II and LAG-3 signaling pathways in Hodgkin lymphoma patients may be a valuable therapeutic target (19). The feasibility of this approach was demonstrated by the ability of IMP321, a soluble LAG-3 fusion protein capable of binding to MHC-II, to inhibit tumor growth in vivo, significantly activating CD8⁺ T cells, and extending PFS in patients with advanced renal cancer (73).

Relatlimab, the first LAG-3 mAb, has shown promising anti-tumor effects in chronic lymphocytic leukemia (CLL) by restoring the immune activities of NK cells and T cells (74) ( Figure 2 ). Combined with lenalidomide (an immunomodulatory drug for multiple myeloma), it enhances IL-2 and NK cell-mediated antibody-dependent cell-mediated cytotoxicity (ADCC) (74). In addition, Thudium et al. reported that another anti-LAG-3 monoclonal antibody (clone C9B7W) effectively inhibited tumor growth in Sa1N fibrosarcoma and MC38 mouse colon adenocarcinoma models (75).

Although targeting LAG-3 alone has shown efficacy in restoring T cell function, many studies indicated combining LAG-3 and other immune checkpoints is more effective. Yang et al. performed a flow cytometric analysis of blood samples from patients with follicular lymphoma (FL) and found that LAG-3 was highly expressed on PD-1⁺ T cells in the TME (76). The ability of PD-1⁺LAG-3⁺ T cells to produce cytokines like IL-2 and IFN-γ, as well as cytotoxic molecules like granzyme B (GzmB) and perforin (PFN), was found to be lower than that of PD-1⁺LAG-3^- cells (76). Woo et al. showed in animal experiments that anti-LAG-3 and anti-PD-1 combination therapy increased the expression of IFN-γ in CD4⁺/CD8⁺ TILs and decreased the level of TNF-α⁺CD4⁺/CD8⁺ TILs (26) ( Figure 3 ). This research demonstrated that combination therapy with anti-LAG-3 and anti-PD-1 slowed cancer progression by restoring and enhancing effector T cell populations in tumors and lymph nodes.

In addition, many studies also support the efficacy of combined blocking strategies in treating different cancer types. Huuhtanen et al. used single-cell RNA sequencing, T cell receptor sequencing (scTCRab-Seq), and other multi-omics techniques to analyze peripheral blood samples from melanoma patients treated with combination therapy with relatlimab and nivolumab (77). According to their research, LAG-3 was highly expressed on CD8⁺ T cells, NK cells, and Tregs in melanoma patients. Moreover, the expression level of LAG-3 was strongly linked to the degranulation activity of NK cells. Following anti-PD-1 and anti-LAG-3 treatments, NK cells degranulated and released cytokines that stimulated the proliferation of CD8⁺ T cells. Additionally, the combination therapy enhanced the cytotoxic properties of antigen-restricted T cells stimulated by NK cells and altered the expression profile of Tregs, compared to LAG-3 blockade therapy alone (77).

Andrews et al. found that CD8⁺ T cells lacking PD-1 and LAG-3 were unique at the transcriptional level in a melanoma mouse model, exhibiting a broad range of TCR clonotypes and abundant effector and interferon response genes, compared with CD8⁺ T cells lacking PD-1 or LAG-3 alone (78). These characteristics allow them to clear tumors more effectively and prolong the life span of mice (78). Ngiow’s study further showed that LAG-3 plays a crucial role in maintaining the persistence of exhausted CD8⁺ T cells (Tex) while generating a subset of CD94/NKG2⁺ Tex cells with enhanced cytotoxicity (41). However, Cillo et al. pointed out that although the combined blockade of LAG-3 and PD-1 can enhance the receptor signaling capacity and cytotoxicity of CD8⁺ T cells, it still retains the exhausted characteristics of Tex, suggesting that further optimization of future treatment strategies still needs to be considered (79).

Kureshi et al. recently examined the effects of LAG-3 and PD-1 inhibition in patients with advanced melanoma (80). They found that the combination therapy improved survival in advanced melanoma patients, with improvements in PFS and OS. Nevertheless, they also observed that patients with combined blockade still exhibited irAEs similar to those with PD-1 blockade alone, possibly due to incomplete blockade or partial overlap in cell types regulated by LAG-3 and PD-1 (80). In the combined blocking experiment of PD-1 and CTLA-4, Larkin et al. found that cancer patients in the nivolumab (PD-1 inhibitor)-plus-ipilimumab (CTLA-4 inhibitor) group showed more severe treatment-related adverse reactions (81). Therefore, the combination therapy with immune checkpoint blockade must be strictly regulated and controlled for different patients.

Collectively, these results demonstrated that inhibiting LAG-3 with other ICs, or the use of neutralizing mAbs against LAG-3 in patients who have become resistant to PD-1/PD-L1 therapy, might be a practical approach for overcoming tumor resistance to mAb therapies and providing therapeutic advantages to patients with acquired resistance. Up till now, LAG-3 blocking has demonstrated synergistic effects in various tumor models when combined with other immune checkpoint therapies, such as anti-PD-1/PD-L1 (26, 75, 82). These promising preclinical outcomes have led to the initiation of clinical trials to assess the safety and efficacy of LAG-3 inhibitors when combined with other ICIs.

Preclinical studies assessed the effect of inhibiting LAG-3 function on tumor progression by blocking the interaction between LAG-3 and its ligand using mAbs. However, the main limitation of mAbs is their difficulty penetrating the dense stroma of tumors, which limits their efficacy and affects clinical outcomes (83, 84). To overcome this challenge, researchers have developed antibody fragments with better penetration, such as single-chain variable fragments (scFv), which may improve clinical outcomes (83, 85) ( Figure 4 ).

Several drug candidates targeting LAG-3 are currently in clinical trials for the treatment of chordoma, esophageal cancer, gastric cancer, and multiple myeloma (86). However, it has been found that these mAbs may cause systemic inhibition of LAG-3 in patients, thereby inducing irAEs, which emphasizes the importance of targeted drug delivery. Targeted delivery of antibodies to tumor sites has been proposed as a potential therapeutic strategy and been mentioned in the treatment of PD-1 and CTLA-4 (87). Based on this strategy, one potential alternative approach is to inject anti-LAG-3 mAbs directly into the tumor with therapies targeting tumor-associated antigens (TAA). This strategy can improve treatment specificity and efficacy, reduce systemic side effects, and maximize therapeutic benefit.

Despite these advances, mAb still faces resistance in some tumors. For instance, in NSCLC or melanoma, some patients do not respond to mAbs treatment or have disease progression after an initial response (88). In response to these problems, small molecule inhibitors are emerging as a promising alternative drug. In contrast to mAbs, small molecules with oral bioavailability enhance tumor penetrating (89, 90). Additionally, a small molecule is more easily optimized for pharmacokinetics, allowing for a flexible dosage regimen to mitigate mAb-related irAEs (91) ( Figure 4 ). The development of LAG-3 small molecule inhibitors may offer a successful immunotherapy for managing certain solid tumors. Abdel-Rahman et al. used a combination of focused screening and “catalog search” to identify small molecules capable of inhibiting LAG-3/MHC-II and LAG-3/FGL-1 interactions (92). Early in vitro tests demonstrated that these small molecules could effectively block LAG-3 ligand interactions, indicating that they may be helpful as therapeutics (92).

Chimeric Antigen Receptor (CAR) T cell therapy represents another promising strategy. Xie and his team validated this therapeutic approach in a mouse tumor model (93) ( Figure 4 ). They grew CAR T cells that recognize the VHH of PD-L1 and applied them to a B16 melanoma mouse model. The results indicated that these organized T cells significantly inhibited tumor progression and improved the survival of the mice (93). Notably, VHH-expressing CAR T cells were specific, active in vitro, and cytotoxic to PD-L1 in melanoma cells, M38 colorectal adenocarcinoma, and HPV16-transformed cell lines, with these properties being mediated through the secretion of IFN-γ. Based on this study, VHH CAR-T cells targeting LAG-3 ligands might be a viable and effective immunotherapeutic strategy. Compared with LAG-3, the ligands such as FGL-1 are expressed at low levels in normal tissues (56). Therefore, targeting ligands can help reduce systemic adverse effects, focus immune regulation on the tumor site, and improve the precision and safety of treatment. In addition, blockade of LAG-3 ligands can inhibit the immunoinhibitory effects of LAG-3 signaling pathway without directly acting on T cells, thereby protecting antitumor T cell activity and enhancing antitumor immune responses in TME.