LILRs are a family of immunomodulatory receptors expressed on myeloid and lymphoid cell lines [14]. These receptors are structurally similar to killer-cell immunolobulin-like receptors (KIRs), whose role in allograft rejection has been highlighted over the past decade [15]. As for KIRs, their encoding genes are located on chromosome 19q [16], and clustered in centromeric and telomeric regions with opposite transcription directions. Depending on the haplotype, LILRA6 and LILRB3 genes have different copy numbers [17, 18]. Within the LILRs’ family, LILRAs and LILRBs subtypes were further defined according to their activators (LILRAs) or inhibitors (LILRBs) functions, the distinction being made according to the activating or inhibitory motifs of their intracellular domain [10]. Both LILRAs and LILRBs are characterized by an extracellular structure comprising several immunoglobulin (Ig) domains. The LILRA-dependent signaling pathways are activated by the recruitment of Src and Syk kinases following the phosphorylation of immunoregulatory tyrosin-based activator motifs (ITAM) sequences present in the intracellular compartment. The Syk and Src kinases phosphorylate PI3K and PLC γ2, leading to activation of the PI3K/Akt, NFAT, Ras/ERK, NF-κB, JNK and MAPK pathways. On the opposite, LILRBs signal transduction include a cytoplasmic domain rich in immunoregulatory tyrosin-based inhibitory motifs (ITIM). After binding of the LILRBs ligands, the phosphorylation of the intracellular ITIM sequences induces the recruitment of the Src homology domain phosphatases SHP-1 or SHP-2. Activation of these phosphatases leads to dephosphorylation of ITAM sequences, particularly those carried by neighboring LILRAs, as well as kinases involved in their downstream signaling. Thereby, LILRB ITIM sequences can interfere with effector functions, cytokine secretion as well as immune cell maturation [19]. For instance, activation of LILRB3 was recently reported to induce NF-κB pathway in cancer cells [20].

LILRB3, also know n as ILT5 (immunoglobulin-like transcript 5), is emerging as a key player in the modulation of the immune response and has attracted growing interest in biomedical research. In the tumor microenvironment, the surface expression of LILRB3 on myeloid cells was recently correlated to a poor patient survival [21, 22]. In addition, a remarkable work reported LILRB3 binding capacity to class I HLA [23], suggesting a new role in regulating immune cells interaction, and not only in regulating immune cell functions. Recently using single-cell RNA sequencing, we showed that LILRB3 expression was increased in kidney allograft infiltrating monocytes during ABMR, suggesting LILRB3 involvement in monocyte activation during rejection [24, 25]. Interestingly the high polymorphism of this gene [26] leads to the translation of slightly different molecules. In 2009, Pfistershammer and colleagues reported two major variants of the LILRB3 protein, ILT5v1 and ILT5v2 based on their respective sequences [27]. Later, Bashirova and colleagues described six different “allotypes” of LILRB3 in the European population. Allotypes 2, 3, 4 and 6 share the same phylogenetic root and correspond to ILT5v1 while allotypes 1 and 5 corresponds to ILT5v2 [26]. These differences may carry the affinity of each variant for a specific ligand. Hence in 2016, Hofer and colleagues investigated LILRB3 variants binding capacity and determined that ILT5v2 bound complement factor C4d and class I HLA unlike ILT5v1 [28] C4d molecules binding to LILRB3 in activated monocytes inhibited TNF-α and IL-6 secretion in a dose dependant manner. Given the major role of complement activation in ABMR [29], the potential interaction of recipient-derived monocyte LILRB3 and C4d in transplant rejection context has yet to be determined.

The advent of in silico tools for modelling protein interactions such as HOMology modeling of COmplex Structure (HOMCOS) [30], allows to speculate that LILRB3 may interact with the heavy chain of antibodies (Fab heavy chain). The ability of LILRB3 to link a DSA Fab, while Fc fragment is bound to another immune cell, is an interesting process to investigate. This interaction could be sufficient to create an immunological synapse between two cells and to inhibit LILR specific pathways (including MAPK, PI3K/Akt or Jak/STAT) leading the monocytes to an anti-inflammatory phenotype. Imbalance of this interaction could switch the phenotype and favor a pro-inflammatory state therefore inducing donor cells destruction.

A large number of proteins have since been reported as potential ligands for LILRB3 including apolipoprotein-E (APOE) [31] and angiopoietin-like (ANGPTL). ANGPTL is involved in diabetes mellitus, in which it promotes the development of adipose tissue, chronic inflammation and systemic insulin resistance in response to T-cell and macrophages activation [32]. In addition, ANGPTL increases expression of TGF-ß and macrophage recruitment, leading to the development and progression of fibrosis in chronic kidney disease [33]. More recently, Galectin 4 and 7 were shown to bind and activate LILRB3 on the surface of myeloid cells and induce an immunosuppressive phenotype [21]. It’s a safe bet that the future will bring the discovery of other LILRB3 ligands, enriching LILRB3 with new functions, not only in the allograft context.

To date, no LILRs have been described as potentially interacting with HLA class II molecules. However, with regard to LILRB3’s ability to bind HLA class I, the literature is contradictory. Indeed, the earliest reports claimed that LILRB3 was unable to bind HLA [34–37] whereas more recent reports [23, 28] contradict this assertion. These discrepancies could be due to the earlier mentioned LILRB3 various allotypes. In silico predictions suggest several potential contact sites for LILRB3 and HLA class I molecules interaction. Of the 4 immunoglobulin domains present in the extracellular part of LILRB3, domains 1 and 2 present binding sites to the invariant HLA class I ß2-microglobulin chain, as well as class I A2 molecule. In addition, domains 3 and 4 appear to have the most interaction sites with HLA I molecules A11, B57, Cw3, Cw4 (Figure 1). LILRB3 polymorphism, especially in domains 3 and 4, could be of utmost importance if balancing the affinity of a particular LILRB3 allotype to an allogeneic ligand, and thus determine the activation status of the recipient myeloid cells toward the donor cells.

Intriguingly at the protein level, LILRB3 shows a strong homology of its extracellular domain with LILRA6, another polymorphic receptor with opposite functions [18, 26, 38]. As expected, this similarity is also found in the gene sequence encoding the two proteins, their homology making them difficult to distinguish either by transcriptomic approaches or using specific antibodies [26]. Unlike LILRB3, several copies of LILRA6 gene can be present in the genome [38]. The homology between these two receptors implies that they bind the same proteins but induce opposite downstream signalization (Figure 2). In basal conditions, a balance exists between the LILRA6 positive signals and the LILRB3 negative signals. Upon strong binding to some LILRB3 allotypes, a disruption of this balance may happen, driving the cell into a specific phenotype. Overexpression of LILRB3 was mostly found on immunosuppressive cells in inflammatory contexts, supporting the hypothesis that LILRB3 induce anti-inflammatory signals in monocytes and leads to pro-resolutive phenotype [21]. With a greater copy number in addition to its high polymorphism, LILRA6 could outperform LILRB3 signals and lead to a pro-inflammatory phenotype. In the future, this sequence homology will have to be taken into account with the utmost vigilance, so that biological functions can be distinctly attributed to LILRB3 or LILRA6.

All known or putative interactions between LILRB3/A6 and their potential ligands of interest, leading to monocyte modulation, are summarized in Figure 2. In the context of kidney transplantation, the recipient-derived monocyte LILRB3 may interact with soluble C4d, directly with donor class I HLA but also with the Fc fragment of HLA-DSA bound to the surface of donor’s renal cells. Other ligands could also be secreted in the synapse for instance ANGPTL, APOE or Galectins, inducing downstream signalization. The ITIM sequences expressed in the LILRB intracellular domain recruits SHP1/2 phosphatases and induce inhibition of central signaling pathways such as PI3K/Akt, Jak/STAT and Pkc dependent pathways. Oppositely LILRA6 intracellular domain interact with ITAM leading to the recruitment of Syk and Src kinases. The recruitment of these proteins activates the same pathways and induce activation, differentiation and proliferation of the monocytes.