Leukocyte Immunoglobulin-Like Receptor (LILR)
The leukocyte immunoglobulin (Ig)-like receptor (LILR) multigene family consists of two pseudogenes (LILRP1 and LILRP2) and 11 functional genes that encode five activating (LILRA1, LILRA2, LILRA4, LILRA5, and LILRA6), five inhibitory (LILRB1, LILRB2, LILRB3, LILRB4, and LILRB5), and one soluble (LILRA3) form. The LILR gene family was independently discovered by different investigators around the same time. Immunoglobulin-like transcripts 1 and 2 (ILT1 and ILT2), which correspond to LILRA2 and LILRB1, respectively, were identified as novel molecules of the immunoglobulin superfamily (IgSF) using IgSF-related PCR primers (Samaridis and Colonna 1997). The leukocyte immunoglobulin-like receptor (LIR-1), which corresponds to LILRB1, was identified as the receptor for the human cytomegalovirus (HCMV) UL18 gene product, an HLA class I mimic (Cosman et al. 1997). MIR-7 and MIR-10, which correspond to LILRB1 and LILRB2, respectively, were identified as gp49B-related molecules (Wagtmann et al. 1997). HM18, HM31/HM43, and HL9, which correspond to LILRB4, LILRA3, and LILRB3, respectively, were identified using a mouse gp49A probe (Arm et al. 1997). CD85 was originally defined at the 5th Workshop on Leukocyte Antigens in 1993 by two monoclonal antibodies, VMP55 and GHI/75, and was shown to correspond to ILT2, owing to the discovery of the LILR genes (Banham et al. 1999); thus, several names (i.e., ILT, LIR, MIR HM/HL, and CD85) have been assigned to this gene family. Although CD85 designations were proposed for this gene family during the 7th HLDA Workshop (Andre et al. 2001), the LILR nomenclature was approved for this gene family by the HUGO gene nomenclature committee (Gray et al. 2015); therefore, this LILR nomenclature is often used in the literature. All of the LILR genes were identified by characterizing their genomic organization (Wende et al. 2000; Wilson et al. 2000).
The LILR family is encoded in the leukocyte receptor complex (LRC) on human chromosome 19q13.4, which contains a number of receptors related to IgSF, such as killer Ig-like receptors. The LRC-syntenic region on mouse chromosome 7 does not encode a direct homolog of the LILR family but, instead, encodes paired immunoglobulin-like receptors (PIR), which have been proposed to be the murine homolog of the LILR family because of its expression pattern, synteny with the LRC region, sequence, and function (Takai 2005). Different numbers of LILR genes are observed among individuals of the same species, indicating that LILR genes are rapidly evolving and are exhibiting greater intraspecies and interspecies differences; this suggests that selective pressures, such as microbial pathogens, could have driven the evolution and diversification of the LILR family (Hirayasu and Arase 2015).
Function in Response to Endogenous Ligands
The signaling molecules downstream of LILRs remain poorly characterized. Inhibitory LILRBs signal through their ITIMs within cytoplasmic domains, and Lck, SHP-1, and Csk contribute to the LILRB1 signaling pathway. In contrast, activating LILRAs signal through ITAMs within the FcRγ chain and Syk kinase is associated with LILRA2 and LILRA4. Structurally, LILRs were initially categorized into group-1 members (LILRB1, LILRB2, LILRA1, LILRA2, and LILRA3) that recognize HLA class I molecules and group-2 members (LILRB3, LILRB4, LILRB5, LILRA4, LILRA5, and LILRA6) that do not recognize HLA class I molecules, according to the LILRB1 residues that interact with its ligand HLA class I molecules; however, this structural categorization is inconsistent with their ligand recognition. LILRA2, a group-1 LILR member, does not bind to HLA class I molecules, although LILRA2 shows >80% sequence identity with LILRB1, and, LILRB5, a group-2 LILR member, binds to HLA class I heavy chains. Crystal structures are available for LILRB1, LILRB2, LILRB4, LILRA2, LILRA3, and LILRA5. The LILR family has unique expression patterns across different tissues, such as hematopoietic lineages and neurons (Fig. 2), and each LILR recognizes different ligands. Thus, LILRs affect diverse biological functions (Burshtyn and Morcos 2016). The following are the functions of individual LILRs and endogenous ligands.
LILRB2. In contrast to LILRB1, LILRB2 recognizes β2m-free forms of HLA class I molecules as well as heavy chain dimers, including HLA-B27 and HLA-G, more strongly than classical HLA class I molecules. Furthermore, LILRB2 binding to HLA class I molecules is affected by peptide antigens presented by HLA class I molecules. LILRB2 also recognizes an HLA-like protein, CD1d, on the cell surface and in the endosomal and lysosomal compartments to block the loading of lipid antigens onto CD1d. Furthermore, LILRB2 recognizes non-HLA ligands, such as angiopoietin-like proteins (ANGPTL), the oligomeric forms of β-amyloid, Nogo66, MAG, and OMgp. LILRB2 is expressed on myeloid cells, hematopoietic stem cells (HSCs), and neuron but not T, B, or NK cells. The murine homolog of LILRB2 appears to be PIR-B, the only inhibitory form of the murine PIR family with respect to expression pattern, ligands, and function; therefore, PIR-B-deficient mice are often used to analyze LILRB2 function. LILRB2 inhibitory functions are dependent on the ligands and cell types. In immune regulation, LILRB2 is involved in the functional inhibition of myelomonocytic cells by recognizing HLA class I molecules. In cell differentiation, ANGPTLs inhibit differentiation and support the expansion of HSCs via LILRB2. In neuronal function, myelin inhibitors, including Nogo66, MAG, and OMgp, are associated with LILRB2 and regulate axon regeneration. In addition, the oligomeric forms of β-amyloid trigger Alzheimer’s disease via LILRB2.
LILRB3. LILRB3 recognizes cytokeratin-8-associated ligands on necrotic glandular epithelial cells, although the function of LILRB3 remains unclear. LILRB3 is expressed mainly on monocytes, DCs, and granulocytes. Allogeneic antibodies against LILRB3 have been detected in HSC transplant recipients and may affect the graft-versus-leukemia effect against LILRB3-expressing leukemic cells. Because LILRB3 is highly polymorphic, the large difference between the amino acid sequences of LILRB3 in donors and recipients might lead to alloantibody production.
LILRB4. LILRB4 is expressed on monocytes, macrophages, DCs, and plasmablast cells. Upregulation of LILRB4 along with LILRB2 is associated with the tolerization of antigen-presenting cells, but its ligands and functions remain unknown.
LILRB5. LILRB5 is expressed on monocytes, DCs, mast cell granules, and T cells and recognizes HLA class I heavy chains, although its function is poorly characterized.
LILRA1. LILRA1 is expressed on monocytes and B cells. LILRA1 has been reported to bind preferentially to the free heavy chains of HLA-B27 and HLA-C but with lower affinity than LILRB1 and LILRB2. Although HLA-B27 was the first ligand identified in the activating LILRAs, it is not known why the activating receptor LILRA1 recognizes HLA class I molecules.
LILRA2. LILRA2 is expressed on monocytes, macrophages, DCs, granulocytes, and minor subsets of T and NK cells. Although its endogenous ligands remain unknown, LILRA2 detects dangerous immunological situations in which antibodies are destroyed by pathogens (as described below).
LILRA3. LILRA3 is the only soluble form of LILR secreted by monocytes, B cells, and subsets of T cells and is not detected on the cell surface. LILRA3 binds to HLA class I molecules with lower affinities than LILRB1 and LILRB2, although the function of these interactions remains unknown. LILRA3 is a high affinity receptor for Nogo66 and promotes synapse formation by competitive blocking Nogo66-mediated inhibition of neurite outgrowth by LILRB2.
LILRA4. LILRA4 expression is uniquely and selectively expressed on plasmacytoid DCs (pDCs) but not on monocytes, T, B, or NK cells. LILRA4 recognizes BST2 on multiple types of cancer cells as a ligand and suppresses pDC activation. Although the exact mechanism of this inhibitory function of LILRA4 through ITAM-containing FcRγ remains unknown, ligand-binding avidity may be partially involved in the function of this inhibitory ITAM. LILRA4 may suppress pDC by recognizing BST2 induced by type I interferons from neighboring cells in a negative-feedback manner.
LILRA5. LILRA5 is expressed on monocytes and neutrophils but its ligands and function are unknown. The crystal structure of LILRA5 was first reported for the group-2 LILR members. LILRA5 shows large changes in structural conformation and charge distribution related to the HLA class I binding sites of LILRB1 and LILRB2, consistent with the observation that LILRA5 does not bind to HLA class I molecules.
LILRA6. LILRA6 expression has been confirmed at the mRNA level in monocytes but not in T, B, or NK cells, and its function remains unclear.
Function in Response to Exogenous Ligands
LILRB1 was identified as a receptor for the HCMV gene product UL18, the first ligand identified in the LILR family (Cosman et al. 1997). UL18 is a glycoprotein homologous to the MHC class I molecules and associates with β2m and endogenous peptides. UL18 binding to LILRB1 was initially thought to be an immune evasion strategy for HCMV to inhibit the immune response through LILRB1 inhibitory signaling (Borges and Cosman 2000; Fig. 3); however, the function of UL18 has remained controversial because UL18 inhibits LILRB1-positive NK cells but activates LILRB1-negative NK cells, as well as LILRB1-positive and CD8-positive T cells. LILRB1 is also exploited by the dengue virus to prevent FcγR-dependent transcription of IFN-stimulated genes and escape from an antiviral response using mimics of the LILRB1 ligand. Bacteria, such as Staphylococcus aureus and Escherichia coli, are also recognized by LILRB1. Mice deficient in PIR-B, a putative LILRB1 counterpart, show resistance to S. aureus, suggesting that LILRB1 may be exploited by bacteria for immune evasion. LILRB2 is also exploited by HIV-1 as HIV-1 mutations at peptides on HLA class I molecules enhance binding to LILRB2 and lead to the functional inhibition of myelomonocytic cells. LILRB3 binds to S. aureus but not to E. coli, although these functional outcomes are unknown.
The LILR family was first discovered in 1997. Since then, a number of studies have suggested a diverse array of LILR functions, including missing self-response, immune tolerance, cell differentiation, targets of microbial immune evasion, antimicrobial response, and nervous system plasticity. In addition, the LILR family is of particular interest for understanding genome evolution because the LILR gene cluster evolved under positive and balancing selection; however, their physiological functions remain largely unknown. The function of LILRB1, which ligands were identified for the first time in the LILR family, remains unknown and controversial because of a lack of animal models to assess physiological roles of specific LILRs. Transgenic animals or nonhuman primate models are necessary to further understand the LILR family. Alternatively, genetic association studies of LILR genes may shed light on the physiological basis of LILR functions in humans, as genome-wide association studies have identified LILR variants that are associated with unexpected phenotypes, such as HDL-C, CK, and LDH. Therefore, the LILR family offers great potential to understand human molecular biology and evolution.
- Hirayasu K, Ohashi J, Tanaka H, Kashiwase K, Ogawa A, Takanashi M, Satake M, Jia GJ, Chimge NO, Sideltseva EW, Tokunaga K, Yabe T. Evidence for natural selection on leukocyte immunoglobulin-like receptors for HLA class I in Northeast Asians. Am J Hum Genet. 2008;82(5):1075–83.PubMedPubMedCentralCrossRefGoogle Scholar
- Hirayasu K, Saito F, Suenaga T, Shida K, Arase N, Oikawa K, Yamaoka T, Murota H, Chibana H, Nakagawa I, Kubori T, Nagai H, Nakamaru Y, Katayama I, Colonna M, Arase H. Microbially cleaved immunoglobulins are sensed by the innate immune receptor LILRA2. Nat Microbiol. 2016;1(6):16054.PubMedCrossRefGoogle Scholar