LAG-3 was discovered in 1990, by Triebel and colleagues, as a new 498-amino acid type I transmembrane protein present on activated natural killer (NK) and T cell lines [
39]. The LAG-3 gene is found close to CD4 on chromosome 12 in humans (chromosome 6 in mice) displaying structural homology to CD4 with extracellular immunoglobulin superfamily (IgSF)-like domains namely D1–D4 [
40]. The structural motifs are conserved between LAG-3 and CD4, translating to the same extracellular folding patterns as a result of which LAG-3 can bind with greater affinity to MHC class II than CD4 [
41]. LAG-3 was speculated to be spatially related to the T-cell receptor TCR: CD3 complex present in microdomains of lipid raft promoting clustering of signaling molecules and the development of the immunological synapse however the exact mechanism is still unclear [
42]. The cytoplasmic tail for the tyrosine kinase p56
Lck, lacks a binding site for LAG-3, which is normally used by CD4 to promote downstream signal transduction of the T cell receptor (TCR) [
41]. Conversely, the LAG-3 cytoplasmic domain has three well-defined motifs namely serine-based motif acting as a PKC substrate, repetitive “EP” motif comprising of a series of glutamic acid-proline dipeptide repeats, and relatively unique “KIEELE” motif, highlighted by an essential lysine residue [
43,
44]. LAG-3 cytoplasmic tailless mutants neither mediate the inhibitory effects of LAG-3 nor compete with CD4, emphasizing the importance of the function of this domain needed for the transmission of an inhibitory signal [
20]. Expression of MHC class II molecules by human melanoma cells is correlated with poor prognosis thus, LAG-3 ligation with MHC-II class, which is seen on melanoma-infiltrating T cells, may facilitate their clonal exhaustion [
45]. In vitro, the demonstration showed that such an interaction may help tumor cells to adopt an escape mechanism giving them protection against apoptosis, with a recent study showing that MHC class II-expressing melanoma cells causes infiltration of tumor-specific CD4
+ T cells, mediated by interaction with LAG-3, which in turn negatively regulates CD8
+ T cell responses [46,47.] Galectin-3 is a ligand, that is expressed by several cells within the TME but not the tumor itself, facilitating interaction with LAG-3 (present on tumor-specific CD8
+ T cells) that may regulate anti-tumor immune activities [
48]. Liver sinusoidal endothelial cell lectin (LSECL) is present in the liver as well as identified in melanoma tumor cells where it stimulates growth by inhibiting anti-tumor T-cell dependent responses [
49]. The interaction between LSECL in melanoma cells and LAG-3 inhibited IFNγ production, mediated by effector T cells (antigen-specific), altering the TME [
49]. Continuous T cell activation in an inflammatory state, specifically in a tumor, results in persistent co-expression of LAG-3 on T cells along with additional inhibitory receptors (IR) such as PD1, TIGIT, TIM3, CD160, 2B4 leading to T cell dysfunction [
50]. Several hematopoietic cell types, including CD11c
low B220
+ PDCA-1
+ plasmacytoid dendritic cells (pDCs) constitutively express LAG-3 [
51] however it is not expressed on any myeloid or lymphoid DC subset.
In vitro, MHC class II-expressing melanoma cells could stimulate LAG-3 positive pDCs to mature and produce IL-6 which was later confirmed in vivo as well with LAG-3 positive pDCs showing increased IL-6 production and an activated phenotype similar to melanoma cells [
52]. Bo Huang et al showed that increased IL-6 promotes the release of CCL2 by monocytes
in vitro, which then may recruit MDSCs thus forming the hypothesis that LAG-3 positive pDCs may indirectly mediate MDSC-related immunosuppression by engaging MHC class II
+ melanoma cells [
53]. LAG-3 functions are regulated by cell surface cleavage mainly ADAM10 and ADAM17 disintegrin /metalloproteases, although in mice soluble LAG-3 seems to have no biological function [
54].