Regulation of T-cell immunity by leucocyte immunoglobulin-like receptors: innate immune receptors for self on antigen-presenting cells

Abstract

Following recognition of microbial patterns, innate immune receptors provide a rapid innate response and trigger antigen-presenting cell maturation to instruct adaptive immune responses. Here we discuss a family of innate immune receptors for self – the leucocyte immunoglobulin-like receptors (LILRs). These LILRs exert powerful inhibitory effects on antigen-presenting cell phenotype and subsequent T-cell responses, and may act to constrain the effects of Toll-like receptor signalling. Despite their broad ligand specificity, differing affinities of LILRs for individual complexes of peptide–major histocompatiblity complex can determine the nature of their effect on downstream immune responses. Expression and function of LILRs may be skewed in certain conditions such as cancer or human immunodeficiency virus infection, particularly by ectopic expression of human leucocyte antigen-G, a high-affinity LILR ligand. We discuss the relevance of LILR-mediated immune regulation across a range of scenarios from autoimmunity to transplant medicine, infection and cancer.

Introduction

An adaptive immune response to infection, which occurs when major histocompatibility complex (MHC) proteins bearing microbial peptides are recognized by T-cell receptors, depends entirely on the action of professional antigen-presenting cells (APCs). But the ability to recognize MHC is not unique to T cells. Professional APCs themselves respond to MHC class I (MHC-I), in this case using members of a family of innate immune receptors known as leucocyte immunoglobulin-like receptors (LILRs). The LILRs act as immunomodulators; their engagement by self proteins can influence APC phenotype, with consequent effects on T-cell stimulation.

Nomenclature of LILRs is complicated because of the simultaneous discovery of these receptors by different laboratories. As a result, they are also commonly referred to as ILTs (immunoglobulin-like transcripts) and LIRs (leucocyte immunoglobulin-like receptors).^1,2 The new standardized LILR nomenclature is the only one to cover all 11 receptors of the family (Table 1). ‘Inhibitory’ LILRs (denoted LILRB) transmit signals through their long cytoplasmic tails, which contain between two and four immunoreceptor tyrosine-based inhibitory domains (ITIMs). Conversely, members of the LILRA subfamily lack their own signalling motifs. Instead they possess a charged arginine residue in their transmembrane domain which associates with the FcεRIγ adaptor protein to transduce signals through its immunoreceptor tyrosine-based activatory domains (ITAM).³

Table 1.

Alternative names, expression profiles and ligands of leucocyte immunoglobulin-like receptor (LILRs)

Principle name	Alternative names	Expression	Ligands
LILRA1	LIR-6, CD85i, LIR6	Macrophages85	MHC-I (HLA-B27 allele)
LILRA2	ILT1, LIR-7, CD85h, LIR7	Monocytes, macrophages, dendritic cells, NK cells, basophils,86 eosinophils87	MHC-I (soluble)
LILRA3	ILT6, LIR-4, CD85e, LIR4, HM43, HM31	Produced by monocytes, only expressed in soluble form	Unknown
LILRA4	ILT7, CD85g	Plasmacytoid dendritic cells5,88	Unknown
LILRA5	ILT11, LIR9, CD85f, LILRB7	CD14+ monocytes89	Unknown
LILRA6	ILT8, CD85b, LILRB6	Unknown	Unknown
LILRB1	ILT2, LIR-1, CD85j, LIR1, MIR7	Monocytes, macrophages, dendritic cells, osteoclasts, eosinophils,87 B cells, T cells, NK cells, placental stromal cells90	HLA-A, HLA-B, HLA-F, HLA-G HCMV UL18 protein
LILRB2	ILT4, LIR-2, CD85d, LIR2, MIR10, MIR-10	Monocytes, macrophages, dendritic cells, osteoclasts, basophils,86 eosinophils,87 placental vascular smooth muscle90	HLA-A, HLA-B, HLA-F, HLA-G, CD1d91
LILRB3	ILT5, LIR-3, CD85a, LIR3, HL9	Monocytes, macrophages, dendritic cells, osteoclasts, basophils,86 eosinophils87	Unknown
LILRB4	ILT3, LIR-5, CD85k, LIR5, HM18	Monocytes, macrophages, dendritic cells, osteoclasts	Unknown, but expressed on activated T cells
LILRB5	LIR-8, CD85c, LIR8	Unknown	Unknown
LILRP1	ILT9, CD85l, LILRA6P	–	–
LILRP2	ILT10, CD85m	–	–

This table lists individual members of the LILR family and their corresponding alternative names. LILRA group members lack cytoplasmic signalling domains of their own, while LILRB carry two to four immunoreceptor tyrosine-based inhibitory domain (ITIM) motifs (see text for details). LILRP refer to non-functional pseudogenes. Current knowledge on the expression profile and ligands for these receptors is also featured.

HCMV, human cytomegalovirus; HLA, human leucocyte antigen; MHC-I, major histocompatibility complex class I; NK, natural killer.

Like Toll-like receptors (TLRs), LILRs are predominantly expressed on myelomonocytic cells and B cells. Their function can also be compared to that of the TLRs: TLRs act as innate immune receptors for microbes and trigger an immune response to non-self, whereas LILRs acting as innate immune receptors for self could provide an inhibitory balancing force. Activation of dendritic cells (DCs) through TLR engagement leads to T-cell priming and an acquired immune response.⁴ In contrast, triggering of LILRs through their interaction with self proteins can modulate DC activation status, antigen-presenting functions and capacity to elicit T-cell responses. The TLR stimulation can influence LILR expression⁵ and in turn LILRs have even been shown to suppress TLR activity.^6,7

Although research into LILR function is in its relative infancy, it is already becoming clear that this family of receptors, with its ability to control immune responses, could have far-reaching effects in an array of immune pathologies. These not only include conditions where an over-exuberant immune response is detrimental to the patient, such as autoimmune diseases, allergies and responses to transplanted organs, but also situations where immune responses are ineffective, such as certain bacterial or viral infections and growth of certain tumours (Table 2). Here we will summarize the recent literature documenting how individual LILR family members on APCs are involved in regulating adaptive immune responses in the context of different disease conditions. Such knowledge will pave the way for future research which focuses on harnessing the immunomodulatory properties of LILRs for therapeutic application.

Table 2.

Known disease associations of leucocyte immunoglobulin-like receptor (LILR) family members

LILR	Known disease associations
LILRA2	Polymorphism associated with susceptibility to autoimmune diseases57
	Up-regulated in lepromatous leprosy lesions: LILRA2 activation suppressed dendritic cell maturation and inhibited the anti-microbial activity of innate immune responses6
LILRA3	Homozygous deficiency associated with increased risk of multiple sclerosis61
LILRB1	Recognizes HCMV homologue of MHC-I1, implicated in immune evasion by HCMV
	Potential use for early diagnosis of HCMV in lung transplant patients50
	Involved in HLA-G-mediated allograft tolerance49,94
	Peripheral blood mononuclear cells from patients with systemic lupus erythematosus exhibit poor LILRB1 function and diminished expression of this receptor on B cells58
	Coexpressed with HLA-G in multiple sclerosis lesions60
	Up-regulated during HIV infection (see text for details)63,64,95,96
	Proportion of LILRB1+ B cells correlates with severity of malaria70
	Involved in HLA-G-mediated tumour immune evasion38,75–77,79
LILRB2	Up-regulated on DCs and ECs in the T-cell suppression cascade which generates allograft tolerance13,23,41
	Involved in HLA-G-mediated allograft tolerance48,94
	Up-regulated in rheumatoid arthritis, suggested role in spondyloarthritis51,53
	Up-regulated by interleukin-10 in HIV patients27
	Higher affinity for HIV-1 viral escape mutant15
	Expressed on B cells in leukaemia patients and tumour cells in lung cancer patients72, 73
	Involved in HLA-G-mediated tumour immune evasion38,75–77,79
LILRB4	Up-regulated on DCs and ECs in the T-cell suppression cascade which generates allograft tolerance13,23,41
	Suppresses pancreatic islet allograft rejection in humanized NOD/SCID mice17
	Role in tumour cell evasion of host immune system18,71
	Expressed on B cells in leukaemia patients72

DCs, dendritic cells; ECs, endothelial cells; HCMV, human cytomegalovirus; HIV, human immunodeficiency virus; HLA, human leucocyte antigen; MHC-I, major histocompatibility complex class I; NOD/SCID, non-obese diabetic/severe combined immunodeficiency.

The influence of LILRs on antigen-presenting function

Dendritic cells, the most potent of APCs, perform dual functions to link innate and adaptive immunity. In response to ‘danger’ signals, they instruct naive T cells to expand into armed effector populations. However, DCs are also responsible for maintaining tolerance to self antigens. The particular role played by DCs hinges on their state of development or activation. Expression and signalling of LILR have been shown to regulate this, and by doing so can prevent effective T-cell stimulation or generate regulatory T cells.⁸

The 11 members of the LILR family differ in their expression profiles and functions. LILRB1 has the widest expression profile and is found on T cells, natural killer (NK) cells, macrophages and DCs. In the context of T-cell responses, it is worth noting that LILRB1 can influence T cells directly through its expression on their surfaces. Subsets of APCs vary in their LILR repertoires. For example, plasmacytoid DCs and myeloid DCs differ in their expression of LILRB4 and LILRA2,⁹ while LILRA4 expression appears to be restricted to only plasmacytoid DCs.⁷ It remains to be seen how the overlapping functions of individual LILRs might contribute to cell phenotype when multiple receptors are engaged.

Recent studies have helped to clarify the function of individual receptors and the effects of their over-expression and/or signalling in certain situations. LILRB1 is up-regulated throughout DC differentiation and its engagement during this process results in a failure of DCs to respond to maturation stimuli.^10–12 DCs that are maintained in an immature state through LILRB1 engagement cannot stimulate effective T-cell responses; this effect seems to be mediated through the interaction between CD80 and cytotoxic T-lymphocyte antigen-4.¹¹ Migration and cytokine secretion profile of DCs are also altered following LILRB1 engagement.^10–12

The work of Suciu-Foca and co-workers has demonstrated the potent tolerogenic effects of two other family members: LILRB2 and LILRB4. Expression of these receptors was strongly enhanced in DCs exposed to alloantigen-specific CD8⁺ CD28⁻ suppressor T (Ts) cells. ‘Tolerogenic’ DCs expressing high levels of LILRB2 and LILRB4 showed an impaired up-regulation of the costimulatory proteins CD80 and CD86, hindering their ability to activate CD4⁺ helper T (Th) cells.^13,14 Such tolerized DCs, and other APCs expressing high levels of LILRB2 and LILRB4, anergize allospecific CD4⁺ CD45RO⁺ CD25⁺ T cells and convert them into regulatory T (Treg) cells, propagating further the T-cell suppression cascade (Fig. 1). The regulatory properties of LILRB2 and LILRB4 may be mediated by different mechanisms. In the case of LILRB2, receptor engagement results in a failure to up-regulate costimulatory proteins (CD80, CD86 and CD40) and MHC.^15,16 In contrast, at least some of the inhibitory effects of LILRB4 may result from ‘reverse signalling’ through its T-cell ligand, as soluble forms of the receptor are sufficient to inhibit T-cell proliferation.^17,18

Figure 1.

Dendritic cells (DCs) are rendered tolerogenic in the T-cell suppression cascade through the up-regulation of leucocyte immunoglobulin-like receptor B2 (LILRB2) and LILRB4. (1) DCs present peptide in the context of major histocompatibility complex class I (MHC-I) to antigen-specific CD8⁺ suppressor T (Ts) cells. (2) CD8⁺ Ts cells interact with DCs, induce expression of LILRB2 and LILRB4 and inhibit CD40-mediated up-regulation of costimulatory molecule expression; DCs become tolerogenic. (3) Tolerogenic DCs fail to stimulate CD4⁺ CD25⁺ T helper (Th) cell proliferation through lack of costimulation. Instead antigen-specific CD4⁺ CD25⁺ Th cells are rendered anergic upon contact with their cognate antigen presented in the context of MHC-II by tolerogenic DCs. (4) CD4⁺ CD25⁺ Th cells differentiate into CD4⁺ CD25⁺ regulatory T (Treg) cells following contact with tolerogenic DCs. (5) Antigen-specific CD4⁺ CD25⁺ Treg cells in turn tolerize other DCs displaying cognate peptide–MHC-II complexes, in a cell contact-dependent manner, by inducing up-regulation of LILRB2 and LILRB4. The cascade continues in this way.

As evidenced by their nomenclature, LILRA2 and LILRA4 signal through ITAM motifs and might therefore be expected to exert ‘activating’ effects on the APC phenotype. Instead, LILRA4 negatively regulates the secretion of type I interferons by plasmacytoid DCs.⁷ LILRA2 also has inhibitory effects on APC function. LILRA2 engagement inhibits DC differentiation, antigen presentation and response to TLR stimulation.¹⁹ Although counterintuitive, signalling through ITAMs to inhibit TLR activity is not a novel concept (reviewed in ref. 20), and MHC-I recognition by ITIM-bearing receptors has also been found to have a stimulating effect on plasmacytoid DCs.²¹

LILR activity both influences and is influenced by cytokine secretion.^19,22 Inhibitory receptors including LILRB1, LILRB2 and LILRB4 can also be up-regulated to generate ‘tolerized’ DC through the action of the cytokines interleukin-10 (IL-10) and interferon-α.^23–27

The nature of self MHC-I as ligands for LILR

Several members of the LILR family have been shown to recognize MHC-I antigen-presenting proteins. The MHC-I are best characterized for their antigen-presenting role to T-cell receptors and, more recently, as ligands for NK cell receptors. The role of MHC-I recognition by cytotoxic cells – to remove cells the MHC-I of which bear viral peptides or are absent – is clear. However, the need for MHC-I recognition by APCs is less obvious, particularly given the broad specificity of receptors such as LILRB1 and LILRB2.

Activity of LILR can be induced through engagement by multiple forms of MHC-I. The best characterized is LILR recognition of MHC-I in trans. But in the case of LILRA2 the only functional ligand documented is soluble MHC-I in serum,²⁸ while LILRB2 has been shown to constitutively associate with MHC-I on the same cell surface in cis.²⁹ It is not clear how competition for engagement by cis, trans and soluble forms of MHC-I might influence LILR activity. A study of murine Ly49 receptors indicated that cis MHC-I binding could mask, and so prevent, a subset of receptors from binding MHC-I on other cells.³⁰ If a similar phenomenon occurs for LILRs then up-regulation of MHC-I on professional APCs could alter the signalling balance between different LILRs in response to MHC-I on other cells.

One aspect of LILR function that has become clearer with recent studies is the importance of ligand affinity and avidity. Despite their broad specificity for all classical and non-classical alleles of MHC-I, signalling through LILRB1 and LILRB2 is increased when they are engaged by higher-affinity ligands and this is reflected in the APC phenotype. Changes as small as a single amino acid escape mutation in a human leucocyte antigen (HLA)-B2705-restricted human immunodeficiency virus (HIV) epitope resulted in a LILRB2-mediated functional impairment of APCs.¹⁵ This raises the possibility that changes in the profile of MHC-bound peptides during disease could alter an individual's immune activation status.

Recent studies have also highlighted the potential of MHC-I dimerization to enhance modulation of APC phenotype by LILRs. HLA-B27 can form disulphide-bonded homodimers through an unpaired cysteine residue in its α1 helix, and these appear to show enhanced binding to LILRB2.³¹ HLA-G, which has a similar unpaired residue in its α1 helix, was also subsequently shown to dimerize through this residue,³² with HLA-G superdimers enhancing the immunosuppressive effects of LILRB1.³³ Recognition of HLA-G in its various forms could be relevant to a range of immune pathologies.

HLA-G as a high-affinity ligand for LILRB1 and LILRB2

Even in the absence of dimerization, HLA-G acts as a high-affinity ligand for LILRB1 and LILRB2³⁴ and expression of this non-classical MHC-I will therefore result in dominant immunosuppressive effects. Multiple isoforms exist including membrane-bound, soluble, disulphide-bonded dimers of heavy chains and disulphide-bonded superdimers of β₂-microglobulin-associated HLA-G (Fig. 2).

Figure 2.

Binding of leucocyte immunoglobulin-like receptor B (LILRB) to human leucocyte antigen G (HLA-G) monomers, superdimers and homodimers. (1) As predicted from the LILRB1-HLA-A2 crystal structure,⁹² the tip of the N-terminal domain, domain 1 (D1), of the LILRB1 molecule binds to the non-polymorphic α3 domain of HLA-G, while the D1–D2 interdomain hinge region binds to β₂-microglobulin (β₂m). Ectopic expression of HLA-G on tumour and pathogen-infected cells can mimic the way in which the semi-allogeneic fetus resists immunological attack from its mother, by interacting with LILR on maternal immune cells. (2) Two HLA-G heterodimers (each comprising heavy chain bound to β₂m) can dimerize through the formation of a disulphide bond between the respective α1 domains, forming a ‘superdimer’ which, by engaging multiple LILR can substantially increase LILRB1-mediated signalling at the cellular level.^33,93 (3) The interaction between D1 of LILRB2 with the α3 domain of HLA-G is stronger and covers a larger surface area compared with that of LILRB1. In fact, whereas LILRB1 depends on β₂m-binding for its recognition of HLA-G, LILRB2 can bind to HLA-G through the D1–α3 interaction alone, in a β₂m-independent manner.⁹³ This differential binding pattern explains why LILRB2 binds strongly to soluble β₂m-free HLA-G5 homodimers synthesized by placental villous cytotrophoblast cells, while LILRB1 cannot. These HLA-G5 homodimers function as effective tolerogenic molecules at the maternal–fetal interface.³⁶

β₂-Microglobulin-associated HLA-G is a high-affinity ligand for LILRB1 and LILRB2, functionally enhanced by disulphide-bonded dimerization. LILRB2, but not LILRB1, recognizes HLA-G free heavy chains,^35,36 increasing its inhibitory effects on costimulation and T-cell responses.³⁷ Ectopic expression of HLA-G itself is sufficient to induce the up-regulation of inhibitory LILR,³⁸ and expression of both HLA-G and inhibitory LILR is up-regulated through the action of IL-10.^25,26,39,40 Ectopic expression of HLA-G as a high-affinity ligand for inhibitory LILR is likely to be an important factor in determining clinical outcomes in transplant tolerance, infection and cancer.

LILRs as immunomodulators in transplantation tolerance

Transplant patients may require a lifelong regimen of immunosuppressive drugs to prevent rejection of their grafted organ. Unfortunately, this also renders them vulnerable to infections and cancer. The APCs are major mediators of allograft rejection because they present MHC alloantigens both directly (intact MHC molecules expressed on donor APC) and indirectly (peptides from allogeneic MHC molecules processed and presented by self APC) to host T cells, bringing about adaptive responses by the host to the donated tissue.⁴¹ Because the LILRs have the capacity to tolerize APCs, they comprise potential therapeutic targets in the pursuit of allograft acceptance based on exploiting natural tolerogenic pathways. This may reduce the requirement for immunosuppression.

In a transplant setting, LILR-mediated triggering of the T-cell suppression cascade could enable tolerance to develop to alloantigens and eventually allografts. This begins with the antigen-specific interaction of DCs with CD8⁺ CD28⁻ Ts cells (which up-regulate LILRB2 and LILRB4 to produce tolerized DCs) and thereafter with CD4⁺ CD25⁺ Treg cells (Fig. 1).¹³ Subsequent waves of regulatory cells perpetuate tolerance by converting emerging generations of effector cells into suppressors.^24,42In vivo studies have shown that CD8⁺ CD28⁻ Ts cells can exert these effects in heart, kidney and liver transplant patients.^13,43,44 Expression of LILRB2 and LILRB4 is also up-regulated on endothelial cells upon interaction with allospecific CD8⁺ Ts cells. The resulting down-regulation of costimulatory proteins endows these non-professional APCs with a tolerogenic phenotype.^41,45

Meanwhile, LILRB1 and LILRB2 may be involved in HLA-G-mediated graft tolerance. Elevated HLA-G levels are associated with better graft acceptance in renal transplant patients.⁴⁶ Work by Naji et al. has shown that LILR recognition of both membrane-bound and soluble HLA-G from the serum of transplant patients was associated with a reduction in T-cell alloproliferation and induction of a regulatory/suppressive T-cell population.⁴⁷ Similar results have been observed using a LILRB2/HLA-G1 transgenic mouse system.⁴⁸ Engagement of LILRB1 by HLA-G also induces the expansion of the CD11b⁺ Gr-1⁺ myeloid-derived suppressor cell population. These are potent natural regulatory cells which promote allograft survival by negatively regulating T-cell function.⁴⁹ Together, these studies suggest the potential to enhance transplant success by exploiting inhibitory LILRs, and in particular their interaction with HLA-G, which could be expressed on the transplanted organ itself.

The LILRs could also have diagnostic applications in transplantation. LILRB1 expression on lymphocytes was shown to correlate with the development of human cytomegalovirus infection in lung transplant patients. The difference in LILRB1 expression between patients with and without the disease was detectable several weeks before detection of the virus itself. Hence, LILRB1 might enable early detection of human cytomegalovirus infection.⁵⁰

LILR expression and function in autoimmune diseases

Abnormal expression of LILRs is associated with serious autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematosus (SLE) (Table 2). LILRB2, LILRB3 and LILRA2 appear to be up-regulated in synovial fluid cells of patients with rheumatoid arthritis, with their levels reduced in response to anti-rheumatic drug treatments.⁵¹ The LILRs have also been suggested to play a role in inflammatory arthritis, partly because of the ability of HLA-B27 to form homodimers with increased LILRB2 binding.^52,53 In addition to altered expression of LILRs in autoimmune disease, it is possible that particular alleles might be associated with certain pathologies. LILRs are polymorphic proteins.^54–56 Studies are now beginning to address the association of LILR alleles with disease. Occurrence of a single nucleotide polymorphism in LILRA2, leading to the expression of a novel isoform on monocytes, was associated with susceptibility to SLE and microscopic polyangiitis.⁵⁷ Peripheral blood mononuclear cells from SLE patients also exhibited poor LILRB1 function and diminished expression of this receptor on B cells.⁵⁸ When DCs from SLE patients were treated with IL-10, enhanced LILRB1 expression was observed compared with controls.⁵⁹ Multiple sclerosis is an autoimmune inflammatory disorder of the central nervous system. LILRB1 and HLA-G, absent from the central nervous system parenchyma under non-pathological conditions, were abundantly coexpressed on macrophages and microglial cells derived from multiple sclerosis lesions.⁶⁰ In addition, LILRA3 deficiency is associated with multiple sclerosis.^61,62

The role of LILR in infectious diseases

Although favourable in transplantation, the inhibitory properties of LILRs may be problematic in certain instances of bacterial or viral infection. Infection with Mycobacterium leprae can result in lepromatous leprosy with high bacterial loads, type 2 cytokine expression and suppression of cell-mediated immunity. LILR family genes, particularly LILRA2, are significantly up-regulated in patients with lepromatous leprosy. In this context, LILRA2 engagement reduced the IL-12/IL-10 ratio towards an anti-inflammatory phenotype and inhibited the antimicrobial activity of the innate immune response to mycobacterial TLR2/1 ligands.⁶ Subsequently, LILRA2 activation was shown to impair immature DC differentiation and antigen presentation to CD1-restricted and MHC-II-restricted T cells.

One infectious disease in which the actions of inhibitory LILRs may be particularly important is acquired immune deficiency syndrome (AIDS). High levels of LILRB1 and LILRB2 are observed in HIV-positive patients, particularly during chronic infection.^15,63,64 This may be the result of elevated levels of inhibitory cytokine IL-10, which up-regulates inhibitory LILRs.²⁷ Increased expression of LILRB2 in monocytes from HIV-positive patients results in a reduced ability to trigger CD4⁺ Th cells,²⁷ an effect that has been shown to be even further increased by viral mutation of an antigenic peptide.⁶⁴ A similar dominant immunosuppressive effect could result from ectopic expression of HLA-G during HIV infection,^65,66 possibly in response to IL-10. The HLA-G polymorphism has been shown to influence the outcome of HIV infection⁶⁷ and unlike classical MHC-I, HLA-G is resistant to surface down-regulation by the HIV nef protein.⁶⁸

Epstein–Barr virus-specific CD8⁺ T cells expressing LILRB1 are found more frequently in individuals who have been infected persistently with the virus for many years; this expression may confer a long-term survival advantage.⁶⁹ Patients with severe malaria presented significantly more LILRB1⁺ CD19⁺ apoptotic B cells compared with those with uncomplicated malaria or healthy controls.⁷⁰

LILR in cancer

There is already a growing literature on the over-expression of these receptors in various malignancies. Soluble LILRB4, which can inhibit T-cell responses and induce CD8⁺ Ts cells, has been observed in patients with melanoma and carcinomas of the colon, rectum and pancreas.^18,71 Although absent from control B cells, LILRB2 and LILRB4 expression was found on B cells from patients with chronic lymphocytic leukaemia.⁷² LILRB2 has also been observed in tumour tissue samples derived from lung cancer patients.⁷³ These findings indicate that inhibitory LILR expression may be a frequent feature of tumorigenesis and, given the ability of LILRs to prevent T-cell priming, may enable tumours to evade a protective immune response. As such, LILRs could serve in the future as targets for intervention in cancer immunotherapy.

In addition to over-expression of LILRs, many tumours exhibit ectopic expression of the high-affinity ligand HLA-G, which could further enhance their ability to prevent T-cell stimulation. Atypical HLA-G expression was first described in 1998 in melanoma cells, when it was identified as a mechanism of tumour escape by means of protection from NK cytolysis.⁷⁴ Since then, HLA-G expression has been shown to be induced in a wide range of malignant lesions (e.g. breast and renal carcinomas, B-cell malignancies), preventing their immune rejection by binding with LILRB1 and LILRB2 expressed on NK cells, cytotoxic T lymphocytes and APCs, consequently inhibiting the anti-tumour activity of these cells.^75–77 In addition to abnormal HLA-G expression on tumour cell surfaces, soluble HLA-G has been found at high concentrations in the sera of certain cancer patients.⁷⁸

The ability of HLA-G to up-regulate LILR expression could further increase immune cell activation thresholds, contributing to tumour escape from immunosurveillance.³⁸ Primary cutaneous T-cell lymphomas co-express HLA-G and LILRB1.⁷⁹ Evidence to date therefore suggests that HLA-G expression impairs immune responses to tumours by engaging with inhibitory LILRs expressed on immune effector cells, transiently blocking their function. As such, HLA-G also represents a promising therapeutic target and could also be exploited for its potential diagnostic and prognostic properties.

Potential for immunotherapy

As the roles of LILRs in various diseases become clearer, their potential as targets for immunotherapy increases. Conceivably, the powerful tolerogenic properties of LILRs could be exploited to dampen unwelcome immune responses in situations such as allograft rejection, allergies and autoimmune diseases. The benefits of antigen-specific immunosuppression in the treatment of these conditions would far outweigh those of systemic immunosuppression, which brings with it increased risk of infections and malignancies. Antigen-specific Ts cells, Treg cells and tolerogenic APCs could be used to create antigen-specific immunosuppression.⁴² For example, pre-treating an organ with an agent to induce expression of LILRB2 and LILRB4 on its APCs could improve tolerance of the graft following transplantation. Conversely, LILR signalling could be blocked with specific antibodies or antagonists to lift the inhibition of T cells and enhance their activity. This would be valuable in cases where stimulating the immune response is beneficial in the treatment of the disease, such as in bacterial/viral infections or cancer.

There are already some drugs that are known to affect LILR function. Modification of HLA-I with the drug amoxicillin hampered its recognition by LILRB1 expressed on NK cells, so lowering NK activation thresholds.⁸⁰ Restoration of T-cell function through the action of amoxicillin might also be predicted, particularly if it blocks interaction with LILRB2. Other pharmacological agents have been shown to induce a tolerogenic phenotype on DCs through the modulation of LILRs. Aspirin-treated DCs displayed enhanced LILRB4 expression and induced hyporesponsiveness in responder T cells.⁸¹ LILRB4 expression on DCs was also enhanced by vitamin D3, although this expression was dispensable for the induction of regulatory T cells.⁸² Vitamin D3, used in combination with dexamethasone, up-regulated DC LILRB2 expression.⁸³ The DC expression of both LILRB2 and LILRB4 was induced following treatment with niflumic acid.⁸⁴ These findings could be instrumental in the development of treatments for autoimmune diseases and allograft rejection.

Conclusion

The increasing amount of literature on LILR function underlines the important roles these receptors play in the control of T-cell activity. The LILRs could prevent overstimulation following TLR engagement and govern the production of regulatory T cells. Future studies should establish how signals from different types of MHC-I (cis, trans and soluble) and their relative levels are interpreted by LILRs on APCs to determine the nature of adaptive immune responses. In a clinical context, there is already evidence to suggest abnormalities in LILR expression during various immune pathologies. This indicates the potential for LILR-targeted immunotherepeutics that could be used in various pathologies – to enhance LILR activity in situations where immune suppression would be beneficial and to block LILR in diseases involving immune evasion.

Glossary

Abbreviations

APC

antigen-presenting cell

DC

dendritic cell

EBV

Epstein–Barr virus

HIV

human immunodeficiency virus

HLA

human leucocyte antigen

ILT

immunoglobulin-like transcript

ITAM

immunoreceptor tyrosine-based activatory domain

ITIM

immunoreceptor tyrosine-based inhibitory domain

LILR

leucocyte immunoglobulin-like receptor

LIR

leucocyte immunoglobulin-like receptor

L-lep

lepromatous leprosy

MHC

major histocompatibility complex

MHC-I

MHC class I

MHC-II

MHC class II

NK

natural killer

pMHC

peptide–MHC complex

RA

rheumatoid arthritis

SLE

systemic lupus erythematosus

TCR

T-cell receptor

Th

CD4⁺ CD25⁺ helper T cell

Treg

CD4⁺ CD25⁺ regulatory T cell

Ts

CD8⁺ CD28⁻ suppressor T cell