Monocytes and macrophages express all members of the LILRB subfamily (32) with LILRB1 being uniformly expressed in monocytes as well as in ex vivo differentiated macrophages (17, 19, 76). LILRB1 cross-linking reduced Ca²⁺ mobilization upon HLA-DR activation (17), and importantly, interfered with activation through stimulation of FcγR (42). Thus, co-ligation of LILRB1 and the activating FcγRI inhibited protein tyrosine kinases and led to reduced tyrosine phosphorylation of the FcRγ-chain and SYK, resulting in decreased intracellular Ca²⁺ mobilization. Therefore, LILRB1 activation was suggested to impair FcγR-mediated effector functions of therapeutic antibodies such as ADCP. Indeed, LILRB1 plays an important role in the regulation of phagocytosis by macrophages (19). The function of LILRB1 as a phagocytosis checkpoint was identified in an attempt to unravel inhibitory pathways that impair the pro-phagocytic activity of the CD47 antibody magrolimab using a panel of solid tumor cell lines (19). LILRB1 gene knockout experiments showed that LILRB1 represents the major receptor for HLA molecules in the regulation of phagocytosis by macrophages.

LILRB1 may also be involved in the regulation of macrophage polarization. Macrophages exist in a range of activation states between the two extremes of classically activated, pro-inflammatory M1 and alternatively activated M2 macrophages (77, 78). A regulatory role was suggested for LILRB1, since knockout of the LILRB1 gene in macrophages in vitro resulted in a higher proportion of macrophages displaying an M1 immunophenotype, possibly by abrogation of interactions of LILRB1 with HLA molecules in cis (19). This function has already been demonstrated for LILRB2. LILRB2 antagonism has been reported to drive tumor-infiltrating myeloid cell polarization toward an inflammatory, M1-like phenotype and to enhance pro-inflammatory and adaptive immune responses (79).

DC differentiated ex vivo from monocytes (moDC) or CD34-positive progenitor cells strongly express LILRB1 (17, 69, 70, 80). During differentiation, LILRB1 was shown to be upregulated and expression was preserved upon DC maturation (70). LILRB1 is also expressed by most human dendritic subsets from the peripheral blood (65). Thus, LILRB1 was found on type 2 conventional DC, plasmacytoid DC and 6-sulfo-LacNAc-positive DC. In these DC subtypes, LILRB1 is upregulated upon toll-like receptor (TLR) stimulation with lipopolysaccharide (LPS) or imiquimod. In contrast, type 1 conventional DC lack LILRB1 expression, even after exposure to TLR ligands. LILRB1 has been suggested to modulate the differentiation of DC and to regulate their functions (70). Thus, DC differentiated from monocytes in the presence of a LILRB1 ligand remained CD14 expression and exerted low HLA-DR levels. The cells did not respond to LPS-induced maturation, did not secrete characteristic cytokines such as interleukin (IL)-10, IL-12 or TGFβ, and exerted a weak T cell stimulating activity (70, 81). Moreover, LILRB1 engagement was shown to inhibit activation by stimulation of the osteoclast-associated receptor (OSCAR), an activating FcR γ- chain associated myeloid receptor involved in antigen presentation (69). Thus, co-ligation of LILRB1 and OSCAR resulted in reduced Ca²⁺ mobilization, impaired cytokine secretion, and a diminished potency to stimulate the proliferation of antigen-specific T cells in vitro (69).

In contrast to monocytes, only a subpopulation of NK cells displays LILRB1 and at lower levels (17). Cell type specific divergent expression was attributed to different promotors in monocytes and NK cells (82). In healthy conditions, LILRB1 expression varies considerably between individuals. The receptor was found on the cell surface of 23% to 77% of NK cells (17), and the observed variation between individuals was linked to polymorphisms in the LILRB1 gene (83). In general, LILRB1 expression differs in NK cell subpopulations and is higher in CD56^dim NK cells, which express FcγRIIIA and are predominant in the peripheral blood, than in CD56^bright NK cells, which lack FcγRIIIA expression (84). In particular, LILRB1 is expressed by terminally differentiated NK cells expressing CD57 and numerous KIR and by virus-induced adaptive NK cells (85–88). Recent findings show that at least under certain conditions LILRB1 contributes to NK cell education - a process that counterbalances the responsiveness of individual NK cells with their sensitivity for inhibition by cognate HLA class I molecules (89, 90). Importantly, LILRB1 engagement by HLA-G impairs the formation of the NK cell activating synapse by inhibiting the polarization of lytic granules and the microtubule organizing centre and impaired filamentous actin accumulation (91). Upon ligation, LILRB1 inhibits Ca²⁺ mobilization, the release of inflammatory cytokines such as interferon (IFN) γ and impairs both natural cytotoxicity and ADCC (17, 91–96).

LILRB1 is also displayed by certain T cells, yet a considerable variation in the frequency of T cells expressing cell surface LILRB1 exists between individuals (97). In healthy donors, LILRB1 is found on the cell surface of about 20% of αβ T cells in younger individuals, but this can increase up to over 50% with age or upon chronic infections (98, 99). LILRB1 surface expression is found predominantly in CD8-positive T cells, in contrast to CD4-positive T cells, which rarely display LILRB1 (98, 100). LILRB1-positive T cells revealed a reduced potential to proliferate but were capable of producing IFN-γ upon T cell receptor (TCR) stimulation (101). Further expression analysis revealed that in CD8-positive T cells LILRB1 was preferentially displayed by the terminally differentiated population of effector memory T cells re-expressing CD45RA (T_EMRA) and marks a T cell population with strong effector functions and a high content of intracytoplasmic perforin (99, 101–103). In addition, cell surface LILRB1 was found also on a subset of effector memory T cells (T_EM). Interestingly, LILRB1 and PD-1 are expressed on distinct T cell populations and induced expression of PD-1 by TCR stimulation was found preferentially in LILRB1-negative T cells in vitro (102).

In experiments with superantigen pulsed antigen presenting cells (APC) LILRB1 was found to colocalize with the T cell receptor (TCR) at the immunological synapse (104). Engagement of LILRB1 interfered with TCR/CD3 signaling and impaired actin cytoskeleton reorganization, cytotoxicity, cytokine production, and attenuated proliferation (17, 93, 94, 104–106). Moreover, LILRB1 was shown to compete with CD8 for binding to HLA class I molecules (53). Thus, in addition to impairing T cell responses by ITIM signaling, LILRB1 may interfere with T cell activation by blocking the binding of CD8. T cells can also acquire LILRB1 from monocytes by trogocytosis (107). The acquired LILRB1 molecules had signaling properties and were shown to regulate T cell functions. LILRB1 is also expressed by subsets of γδ T cells and contributes to their regulation (17, 108). For example, gene knock-down of β2m or antibody blockade of LILRB1 enhanced lysis of lymphoma cells by γδ T cell clones (108).

LILRB1 expression in B cells increases throughout maturation in the bone marrow and is maintained at the plasma cell stage (16, 82). In peripheral blood, mature B cell populations including naïve and memory B cells homogenously express LILRB1, whereas only a fraction of transitional B cells were stained positive for LILRB1 (82). LILRB1 has been suggested to participate in the regulation of B cell functions. Thus, antibody-mediated crosslinking of LILRB1 interfered with B cell receptor activation resulting in reduced Ca²⁺ mobilization (17). Subsequent findings indicated that LILRB1 activation dampened IgG production, inhibited isotype switching and impaired cytokine production (71). In addition, HLA-G aggregated on nanoparticles interfered with B cell responses by binding to LILRB1 (72). In these experiments, inhibition of T cell-dependent as well as T cell-independent B cell responses was observed and aggregated HLA-G reduced B cell proliferation, antibody secretion and cytokine release.

Alterations in HLA expression that emerge during the evolution of immune escape variants are frequently found in tumors (109, 110). Down-modulation of classical HLA molecules enables tumor cells to evade from attack by T cells, which typically recognize peptides presented on HLA molecules. In contrast, HLA class I molecules function as markers of self and inhibit NK cell functions (11). As a consequence, reduced or loss of HLA class I expression renders tumor cells susceptible to NK cell cytotoxicity. Typically, HLA class I recognition involves KIR and NKG2A receptors, but also LILRB1 plays a role. Thus, masking of LILRB1 enhanced NK cell cytotoxicity against HLA class I-expressing leukemia cells, in particular when KIR or NKG2A were blocked simultaneously (111). LILRB1 was suggested to interfere with activation of NK cells upon engagement of natural killer group 2 member D (NKG2D), an important stimulatory NK cell receptor that regulates natural cytotoxicity, and to inhibit lysis of leukemic cells with ectopic expression of the NKG2D ligand MHC class I chain-related protein A (MIC A) (92).

Intriguingly, HLA class I molecules not only inhibit NK cell-mediated lysis of tumor cells but also function as ‘Don`t Eat Me!’ signals that prevent malignant cells from phagocytic uptake by macrophages through LILRB1 engagement (19). In the tumor microenvironment, tumor-associated macrophages (TAM) contribute to cancer progression by supporting cancer cell survival and proliferation, promotion of angiogenesis, and suppression of immune responses (77, 78, 112). Yet, macrophages are also able to eliminate malignant cells by phagocytosis and to promote anti-tumoral functions of lymphocytes. The role of MHC class I in the regulation of anti-tumoral responses by macrophages was demonstrated in xenograft studies using genetically modified human tumor cells expressing a mouse human chimeric β2m protein to confer interaction with murine macrophages (19). These experiments revealed that MHC class I protected tumor cells from macrophage attack in vivo. Genetic deletion of HLA class I cell surface molecules augmented ADCP by anti-epithelial cell adhesion molecule (EpCAM) or anti-epidermal growth factor receptor (EGFR) antibodies, suggesting that the HLA class I:LILRB1 axis may compromise the therapeutic efficacy of tumor targeting antibodies. Thus, loss of HLA class I may sensitize tumor cells to phagocytosis.

Tumors may also impair immune cells by neo-expression of non-classical HLA molecules such as HLA-G. HLA-G expression was described in various tumors including gastric, colorectal, lung, breast, hepatocellular and pancreatic cancer as well as chronic lymphocytic leukemia and was associated with tumor progression and a worse outcome (113–116). Both cell membrane-bound HLA-G and its soluble form exert pleiotropic immune suppressive functions on various immune cells including T cells, NK cells and APC, which involve LILRB1 signaling (114). For instance, in NK cells, interaction of HLA-G and LILRB1 was shown to counteract signaling and induction of cytotoxicity by interaction of the activating NK cell receptor NKG2D with its ligand MIC A on tumor cells (117). Interestingly, HLA-G upregulates the expression of LILRB1 in NK cells, T cells and APC in vitro (118), which may further potentiate immune suppression.

The analysis of cells in the tumor microenvironment of different types of solid tumors confirmed that LILRB1 was mainly expressed in the tumor stroma with TAM representing the major LILRB1 expressing immune cell population (113). Of note, LILRB1 was found to be displayed by TAM from different types of solid tumors including colon carcinoma, head and neck cancer, non–small cell lung cancer, renal cell carcinoma and by lymphoma associated macrophages (19, 76, 113). The analysis of the tumor microenvironment in gastric cancer revealed that LILRB1-positive TAM had an M2-like phenotype. Their numbers correlated with the levels of immune suppressing cytokines and were associated with T cell exhaustion and increased expression of inhibitory immune checkpoints such as PD-1 and CTLA-4 (119). Beside TAM, LILRB1 was expressed by a significant number of NK cells (113). In certain tumors, the fraction of LILRB1 expressing NK cells was increased in the tumor microenvironment. For example, in prostate cancer, a higher percentage of tumor infiltrating NK cells expressed LILRB1 in comparison to NK cells from control tissues (120). Tumor-induced upregulation of LILRB1 was supported further by subsequent findings revealing that certain tumor cells, i.e. prostate cancer cells and glioblastoma cells, are capable of inducing the expression of LILRB1 on NK cells in vitro (120, 121). In addition, cell-to-cell contact with M2 polarized macrophages augmented LILRB1 expression in NK cells, which led to reduced cytotoxicity and cytokine production (122). In contrast, LILRB1 was more rarely expressed in tumor infiltrating T cells and rather restricted to CD8-positive T cells with a T_EMRA phenotype. Of note, in the peripheral blood of cancer patients an increased frequency of LILRB1 expressing NK and T cells was observed compared to healthy controls (92, 96, 113, 120, 123, 124).

The expression of LILRB1 in the tumor microenvironment may have prognostic features. In gastric cancer for example, elevated LILRB1 expression was associated with advanced tumor stages, increased risk of recurrence and inferior survival (119). In prostate cancer, enhanced levels of LILRB1 mRNA were associated with a shorter biochemical recurrence-free survival according to prostate specific antigen levels in the blood (125). Moreover, in ovarian cancer a high content of LILRB1-positive immune cells was correlated with shorter survival, and worse adjuvant chemotherapy responses (126).

LILRB1 is also expressed by malignant cells of different tumor types (43). Thus, LILRB1 expression was reported in AML (127), certain T cell lymphomas (128–130), B-lineage lymphomas and leukemias (131–133), gastric cancer (134), ovarian cancer (126), and renal cell carcinoma (135). LILRB1 expression by tumor cells may have different consequences. For example, LILRB1 expression protected cutaneous T-cell lymphoma cells from CD3/TCR activation induced cell death (128), while in malignant B cells LILRB1 engagement dampened tumor cell proliferation and induced cell cycle blockade (132). Intriguingly, the expression of LILRB1 may increase the susceptibility of cancer cells to an attack by immune cells. In multiple myeloma, loss of LILRB1 has been suggested to contribute to immune escape by reducing their susceptibility to NK cell mediated lysis (133). The LILRB1 ligand S100A9 expressed by NK cells was suggested to play a role and effects of LILRB1 expression on cytotoxicity were diminished by S100A9 blockade. Moreover, LILRB1 expressed by lymphoma cells sensitized these cells to lysis by certain γδ T cells, presumably by co-stimulation of the effector cells through ligation of HLA class I on γδ T cells (131).

In initial experiments, Barkal and colleagues employed the murine hybridoma antibody clone GHI/75 to disrupt the HLA class I:LILRB1 axis (19). Similar to HLA class I-specific fragments antigen binding (Fab), this anti-LILRB1 antibody enhanced the phagocytosis of HLA class I-expressing solid tumor cell lines in vitro, but required combination with the CD47 directed IgG4 antibody magrolimab to become effective (19). Notably, CD47 IgG4 antibodies may exert dual function by blocking CD47 at the tumor cell site and engaging activating FcγR at the effector cell side to provide an ‘Eat Me!’ signal (9, 19, 142). Regarding the suggested interplay between the two axes CD47:SIRPα and HLA class I:LILRB1 in counteracting effective ADCP of tumor cells as demonstrated for anti-EpCAM or anti-EGFR antibodies (19), additional antibody combinations were studied. For example, a chimeric, Fc-silent variant of antibody GHI/75 with abrogated FcγR binding was generated, to prevent potential occurring effects by binding of this antibody to FcγR (76). In combination with a CD20 antibody (i.e. rituximab or obinutuzumab) for FcγR engagement and an Fc-silent CD47 antibody to inhibit SIRPα signaling, LILRB1 antibody blockade further enhanced the phagocytic uptake of lymphoma cells by macrophages in vitro compared to the combination of CD20 antibodies and CD47 blockade only. This was observed in studies with non-polarized macrophages as well as macrophages that were polarized towards M1 or M2 phenotypes ex vivo, and lymphoma-associated macrophages isolated from the bone marrow of lymphoma patients. The blockade of LILRB1 alone however was not sufficient to induce phagocytosis in this study. In addition, the LILRB1 antibody was not effective when combined only with either a CD20 or a CD47 antibody. These results suggested that both FcγR engagement by the tumor targeting antibody and blockade of the CD47:SIRPα axis were required (76). In contrast, characterization of the humanized IgG4 anti-LILRB1 antibody BND-22 revealed that this antibody was effective also as single agent in inducing phagocytosis of tumor cells by human macrophages both in vitro and in a xenograft in vivo model in which both human macrophages and cancer cells were transferred (113). In addition, the antibody enhanced ADCP by the anti-EGFR antibody cetuximab in vitro, thereby providing further evidence that LILRB1 blockade may enhance the therapeutic efficacy of tumor targeting antibodies by enhancing their ADCP function.

Regarding the tumor promoting and immune suppressive functions of TAM, reprogramming TAM to a pro-inflammatory state may inhibit tumor growth. In fact, LILRB2 antagonistic antibodies were shown to exert this immunomodulatory function (79). Yet, whether LILRB1 blockade allows promoting polarization towards an inflammatory M1-like phenotype and whether this may offer the opportunity to also reprogram TAM, has not been addressed experimentally to our knowledge and will require further investigation ( Figure 2A ).

NK cells play an important role in immunosurveillance against cancer due to their natural cytotoxic functions and represent an important effector cell population for therapeutic antibodies as mediators of ADCC. LILRB1 blockade may allow both promoting natural cytotoxicity and enhancing ADCC ( Figure 2B ). For example, antibody masking of LILRB1 was demonstrated to promote the activation of NK cells from chronic lymphocytic leukemia (CLL) patients, to enhance NK cell proliferation when combined with the immunomodulatory drug lenalidomide and to promote lysis of CLL cells (96). Similarly, enhanced NK cell cytotoxicity against glioblastoma cells was observed in vitro when LILRB1 was blocked (121). A humanized, Fc-engineered anti-LILRB1 IgG antibody, B1-176, with reduced FcγR binding was shown to disrupt LILRB1 signaling in co-culture experiments with LILRB1 reporter cells and various tumor cell lines (92). The antibody enhanced natural cytotoxicity by the NK cell line NKL against HLA class I-positive leukemia cells and particularly supported lysis induced by NKG2D engagement. In addition, antibody B1-176 promoted anti-tumoral activities of NK cells in ALL or multiple myeloma xenograft experiments in vivo. Regarding ADCC, LILRB1 was shown to hamper cetuximab-mediated ADCC by NK cells. Thus, antagonistic LILRB1 antibodies restored the cytotoxic activity of NK cells from triple-negative breast cancer patients and enhanced ADCC (123).

T cells are an attractive effector cell population in cancer immunotherapy, in which T cell checkpoint inhibitors or bispecific T cell engager (BiTE) molecules have shown promising results (3, 4). The ability of anti-LILRB1 antibodies to enhance T cell activation was demonstrated in mixed lymphocyte reactions with allogeneic T cells and DC, in which the antibody BND-22 enhanced T cell activation as verified by IFN-γ release (113). Recent findings suggest that LILRB1 engagement may impair the efficacy of both BiTE and PD-1 immune checkpoint inhibitors ( Figure 2C ). LILRB1-expressing T_EMRA cells represent an important T cell population for CD3 bispecific antibodies with a high cytotoxicity (102). The anti-LILRB1 antibody GHI/75 has shown potential to enhance T cell cytotoxic activity mediated by an [anti-MART-1 × CD3] BiTE molecule by disrupting the interaction of LILRB1 with HLA-G expressed by malignant melanoma cells (102). Regarding PD-1 immune checkpoint blockade, LILRB1 upregulation has been discussed to be a potential mechanism of resistance. Thus, increased expression of LILRB1 was reported in nearly half of malignant melanoma patients after treatment with the PD-1 directed antibody nivolumab (113). Interestingly, combining anti-PD-1 and anti-LILRB1 antibodies synergistically enhanced the secretion of TNF-α by autologous peripheral blood mononuclear cells co-incubated with colon cancer cells in vitro and improved therapeutic effects were observed for the antibody combination in vivo (113). Moreover, since PD-1 and LILRB1 are described to be expressed by different T cell subsets, co-blockade of both receptors may provide an opportunity to abrogate inhibitory signaling pathways in both subsets (102).