As a novel therapeutic approach, immune checkpoint blocking therapy, which can reactivate T cell immune responses to tumor cells and break tumor immune suppression, has achieved marked success in preclinical or clinical trials of many malignant tumors(5, 7, 8, 11, 40), and the most extensively used immune checkpoint inhibitors for the study and application of cancer therapy include PD-1 and inhibitors of its ligand PD-L1, as well as CTLA-4. However, the objective response rate has been 13%~56% and the complete response rate has been only 1%~16%, which presents a frustrating challenge, especially for breast cancer (5–11). Therefore, it is urgent to make further progress on the tumor microenvironment to explore alternative or facilitating therapeutic targets. Many recent studies have shown a special correlation between LAG3 and PD-1 of T cell inhibition in various diseases (12, 41), such as in viral infection (12, 42), chronic tuberculosis infection (43), plasmodium infection (44), chronic lymphocytic leukaemia (45), ovarian cancer (46), and so forth, yet the co-expression and effect of LAG3 and PD-1 on T cells in breast cancer patients are still not very clear. To figure out the molecular and clinical relationship between LAG3 expression and immune activities in breast cancer will greatly promote the establishment and progress of a novel therapeutic target as well as optimize the current therapeutic strategies.
In this study, we systematically analyzed the expression of LAG3 in breast cancer. We found that LAG3 was upregulated in breast cancer tissue, especially enriched in basal, HER-2 enriched and LumA subtype, as well as in patients with higher tumor grade. LAG3 also had potential to serve as valuable biomarker for TNBC subtype. Besides, previous studies also found that the presence of LAG3 + intra-epithelial tumor infiltrating lymphocytes (iTILs) was significantly related to younger age, large tumor size, ER/PR negativity and high Ki67 proliferation index (47, 48). All these results seemed to indicate that the high expression of LAG3 predicted a high malignant breast cancer. However, some studies had come to the seemingly opposite conclusion that high expression of LAG3 is associated with favorable overall survival in several solid tumors including ovarian, gastric, lymphoma, NSCLC, colorectal, and renal cancers (49), as well as in breast cancer. To make a reasonable explanation, we need to focus on the important immunologic role of LAG3.
Under physiological conditions, LAG3 is expressed on membrane of activated human T cells, NK cells, B cells and DCs (50–53), and it is an activation marker for CD4 + and CD8 + T cells. In tumor patients, LAG3 is expressed on surface of TILs (54, 55). Early studies suggested that LAG3 was a negative regulator of T-cell activation, and the regulation of T cells immune response mainly by following three aspects. Firstly, the proliferation and activation of T cells were inhibited directly by negative regulation. Previous studies showed that LAG3 was a negative regulator of T-cell activation, and blockade of LAG3 functioning on human CD4 clones resulted in enhanced cell proliferation with an elevated production of IFN-γ, TNFα, IL- 2 and IL-4 (16). Furthermore, there was a largely conserved motif named KIEELE mediated a cell intrinsic signal, which is thought to be essential for the negative regulatory function of LAG3 on T cells (56). A more specific role for LAG3 on CD8 + T cells was demonstrated in a model of self-tolerance. Grosso JF et al. found that adoptively transferred LAG3−/− HA-specific CD8 + T cells expanded and produced large amounts of IFNγ, indicating that LAG3 limited self-tolerance (52). Moreover, they also found that these CD8 + T cells regained effector function, as was shown with an increased number of IFNγ-producing cells. Therefore, hypothesis was that not dependent on CD4 + T cells, the effect induced by blocking LAG3 was shown to be a CD8 + T-cell intrinsic effect (52). Secondly, T cell immune response was suppressed indirectly by promoting inhibitory function of regulatory T cells (Treg). Recent studies have shown that LAG3 promoted the differentiation of Tregs, while its blockade inhibited the induction of Tregs (57). This study also illustrated that CD4 + T cells were skewed into a Th1 phenotype by blockade or genetic deletion of LAG3, with LAG3 limiting IL-2 and STAT5 signaling that modulated the ability to be suppressed by Tregs. Thirdly, T cell activation was prevented by regulating antigen presenting cells (APC) (14), which was also supported by our finding that LAG3 had close correlation with antigen processing and presentation pathways. Published studies also showed that LAG3 may be invoved in mediating bidirectional signaling into the interacting APCs. DCs activation was shown to be inhibited by MHC class II binding to LAG3-expressing Tregs, thereby their maturation was suppressed (58).Interestingly, previous studies have been focused on the impact of LAG3 on T cell immunity, whether LAG3 have impact on other immune response and immune cell populations was unclear. Herein, we found that LAG3 was positively correlated with B cell mediated immunity, B cell activation, B cell receptor signaling pathway and natural killer cell mediated cytotoxicity pathways. Consistent to our observations, previous studies indicated that LAG3 expression was also related to NK cells and activated B cells in a T cell-dependent manner (59).
Besides, in terms of immune cells, we observed LAG3 expression had strongest correlation with T cells (especially CD8 + T cells), followed by plasmacytoid dendritic cells, NK cells, monocytic lineage and B lineage. In this section, as the relationship between LAG3 and T cells was already mentioned above, the correlation with other immune cells is focused. LAG3 is constitutively expressed on plasmacytoid dendritic cells (pDCs) at a much greater level than any other cell type (60), while LAG3 is not expressed on any lymphoid DC and myeloid subset. When compared with wildtype pDCs, LAG3- pDCs showed enhanced expansion following CpG stimulation in vivo, but had no altered expression profile of activation markers, including differential cytokine production or CD80/86 and MHC class II (60). As was expounded in earlier studies, in humans, LAG3 + pDCs were found to involved in the melanoma environment and interacted with HLA-DR-expressing tumor cells in vivo. Besides, in vitro it was also shown that as the result of the stimulation of MHC class II-expressing melanoma cells, LAG3 + pDCs were able to mature and produce IL-6 (61). Therefore, there was a hypothesis that LAG3 + pDCs may indirectly drive myeloid-derived suppressor cells (MDSCs)-mediated immunosuppression through MHC class II + melanoma cells engagement. LAG3 is also found expressed on NK cells (~ 10%) and invariant NKT cells (19). As a result of LAG3 signaling, proliferation of activated NKT cells was reduced, resulting in cell cycle arrest in the phase S (62). Moreover, researchers also found that overexpression of LAG3 was associated with impaired iNKT cytokine production (IFNγ) during chronic HIV infection, nevertheless this was not discovered on other T-cell subsets (63). As for monocytic lineage, one previous study had suggested that a soluble monomeric form of LAG3 (sLAG3), generated by alternative splicing, impaired the differentiation of monocytes into DCs and macrophages, which subsequently had diminished its immunostimulatory capacity (64). Moreover, another previous study had reported that at the end of IMP321 (a LAG3 antagonist) treatment, there was a 50% objective tumor response and decreased tumor size related with an increase in the absolute number of monocytic cells (65). The role of LAG3 on B cells is controversial since analysis was limited and expression was only reported in a single study (44). In short, LAG3 exerts differential inhibitory impacts on various types of lymphocytes. Except for some relatively deep and detailed researches towards T cells, the functional role and mechanism of LAG3 on other immune cells are not fully understood and well established, and further studies are needed to enrich this filed.
Back to the question posed at the beginning of this section, LAG3 expression was found to be associated with poor clinicopathological factors and eliciting immune suppressive function, supporting the hypothesis that the expression of LAG3 in breast cancer patients should lead to poor survival. However, inconsistent with the present results, the findings of several previous studies indicated a favorable association between high expression of LAG3 and cancer-specific survival, especially in ER-, HER2 + and basal-like subtypes (47, 66, 67). Interestingly, another published study found that serum LAG3 had a close correlation with prolonged survival in ER + patients (42). These results implied a complicated relationship between LAG3 expression, clinical characteristics and the prognosis in breast cancer. One possible explanation is that presence of LAG-3 expressing TILs may actually indicate that there is an ongoing cancer-immune interaction (47), a phenotype defined as an inflamed tumor (68), and it usually signified a somewhat improved prognosis. Additionally, another study also illustrated that LAG3 expression by engineered tumor cells efficiently promotes and facilities activation, intra-tumoral recruitment, and Th1 commitment of APCs, which results in a large intra-tumoral influx of specific and non-specific reactive cells, as well as the release of immunoregulatory and cytotoxic mediators (69). Consequently, further studies are encouraged and needed to focus on this controversial problem.
Despite the promising impact of cancer immunotherapy targeting CTLA4 (like ipilimumab) and PD1/PDL1 (like pembrolizumab), with the in-depth research, the side effects and resistance of these drugs have gradually emerged (70, 71). Moreover, a large number of cancer patients fail to respond: the response rate of ipilimumab was only 15%, and that of PD-1/PD-L1 inhibitors was less than 40% (72). In this study, we found that LAG3 had close correlation with PD-L1 expression and PD-1 checkpoint pathway in cancer, and strong correlations were observed between LAG3 and other checkpoint members, such as CTLA4, TIGIT, CD28, CD40, CD48 and other checkpoint molecules including CD27, CD86, ICOS, IDO1 and so on. Early studies have reported that sustained T-cell activation induced by a chronic inflammatory environment, for example, during chronic viral infection or in a tumor, caused persistent LAG3 expression on T cells which frequently co-express with other IRs, like PD1, TIM3, TIGIT, CD160, 2B4 and so on, subsequently resulting in a T-cell dysfunction state (73). This state, also named T-cell functional exhaustion, defined by a distinct subset of exhausted T cells with an elevated expression of IRs, resulting in lack of proliferation, cytokine secretion, and cytolytic activity (15, 74–76).