Molecular recognition of paired receptors in the immune system

Abstract

Cell surface receptors are responsible for regulating cellular function on the front line, the cell membrane. Interestingly, accumulating evidence clearly reveals that the members of cell surface receptor families have very similar extracellular ligand-binding regions but opposite signaling systems, either inhibitory or stimulatory. These receptors are designated as paired receptors. Paired receptors often recognize not only physiological ligands but also non-self ligands, such as viral and bacterial products, to fight infections. In this review, we introduce several representative examples of paired receptors, focusing on two major structural superfamilies, the immunoglobulin-like and the C-type lectin-like receptors, and explain how these receptors distinguish self and non-self ligands to maintain homeostasis in the immune system. We further discuss the evolutionary aspects of these receptors as well as the potential drug targets for regulating diseases.

Keywords: ITAM; ITIM; c-type lectin-like receptor; immunoglobulin-like receptor; infectious diseases; paired receptor; structural biology; tumorigenesis.

Figure 1

(A) Schematic representation of the LRC on human chromosome 19q13.4. A large number of Ig-like receptor genes, including two clusters of LILR loci and a cluster of KIR loci, are encoded within the LRC. Arrows indicate the direction of transcription for each gene. These loci have evolved by multiple duplications, and the two LILR clusters are likely to have been generated by the inverse duplication of an ancient one. (B) The hypothesis of paired receptor family evolution. I) The NK cell possesses at least one inhibitory receptor (cyan), and the inhibitory signals through it protect the normal self-cells from NK cell killing. II) In the infected cells, the low level expression of MHCIs induces the NK killing, in a system called the “missing-self hypothesis.” III) In order to escape the host NK cytotoxicity, some viruses acquired the expression of MHCI-like molecules (purple), which bind to the inhibitory receptors. IV) On the other hand, NK cells express activation receptors (orange), which evolved from the related inhibitory receptors to trigger NK cell activation.

Figure 2

Domain configuration of the KIRs. The extracellular Ig-like domain is classified into three types, D0, D1, and D2, dependent on the sequence homology. KIR2DL4 possesses an ITIM motif, but also associates with the FcRγ chain.

Figure 3

Comparison of the recognition modes of the MHCI/MHCI receptors. The complex structures of HLA-A2 and TCR (red; A, PDB ID:2VLR), HLA-Cw4 and KIR2DL1 (orange; B, PDB ID:1IM9), and HLA-G and LILRB2 (yellow; C, PDB ID: 2DYP). The MHCIs (heavy chain in green, β2m in cyan, peptide in magenta) are recognized in different manners. (A) TCR binds to the center of the peptide and the α1–α3 domain of HLA-A2. (B) KIR2DL1 binds to both the α1 and α2 helices of HLA and the C-terminal end of the peptide. This binding region contains the 77N/S and 80K/N residues, which determine the ligand specificity. (C) LILRB2 binds to the α3 domain and β2m, which are conserved regions among the MHCIs.

Figure 4

Structure of UL18/LILRB1 and comparison with the HLA-A2/LILRB1 complex. (A) The crystal structure of the UL18/LILRB1 complex (UL18 in green, β2m in cyan, peptide in magenta, LILRB1 in yellow; PDB ID: 3D2U) with complex carbohydrate models attached to the 13 potential N-glycosylation sites. The α1–α2 domains recognized by TCR or KIRs are highly glycosylated, and steric hindrance inhibits effective interactions. Meanwhile, the interface with LILRB1 still is exposed. (B) The crystal structure of the HLA-A2/LILRB1 complex (HLA-A2 and peptide in purple, β2m in cyan, LILRB1 in yellow; PDB ID: 1P7Q).

Figure 5

The complex structure of NKG2A/CD94 and HLA-E. (A) The overall structure of the NKG2A/CD94/HLA-E complex (PDB ID: 3CDG). NKG2A/CD94 recognizes the α1–α2 domain of HLA-E containing a peptide. (B) The interface of NKG2A/CD94 and HLA-E. The residues interacting with P5Arg of the peptide are depicted by red stick models, and P8Phe is shown by a blue stick model.

Figure 6

Schematic model of LLT1 recognition by NKRP1 and comparison of the structures of the Ly49-MHCI complex and m157. (A) The model structures of LLT1 (cyan) and NKRP1 (blue) are shown by ribbon models. Residues that may contribute to the interaction with LLT1 are shown as spheres, with detrimental effects in red, and modest effects in orange. Magenta spheres indicate the pair of residues that showed detrimental effects when mutated independently, but restored the binding when mutated simultaneously (Kamishikiryo et al., 2011). (B) The crystal structure of Ly49C with H-2K^b (H-2K^b in green, β2m in cyan, Ly49C in blue; PDB ID: 1P4L). Ly49C interacts with the β2m subunit and the α3 domain of H-2Kb, but not with the peptide-binding region. The crystal structure of m157 (PDB ID; 2NYK). (C) The mutation sites that identified the Ly49H binding residues are mapped and are shown in stick style (red). The residues did not overlap with the interface of the Ly49-MHCI complex structure.