The SLAM family receptors: potential therapeutic targets for inflammatory and autoimmune diseases

Abstract

The signaling lymphocytic activation molecule (SLAM) family is comprised of nine distinct receptors (SLAMF1 through SLAMF9) that are expressed on hematopoietic cells. All of these receptors, with the exception of SLAMF4, are homotypic by nature as downstream signaling occurs when hematopoietic cells that express the same SLAM receptor interact. The SLAM family receptor function is largely controlled via SLAM associated protein (SAP) family adaptors. The SAP family adaptors consist of SAP, Ewing sarcoma associated transcript (EAT)-2, and EAT-2-related transducer (ERT). These adaptors associate with the cytoplasmic domain of the SLAM family receptors through phosphorylated tyrosines. Defects in SLAM family members and SAP adaptors have been implicated in causing immune deficiencies. This is exemplified in patients with X-linked lymphoproliferative (XLP) disease, where SAP undergoes a loss of function mutation. Furthermore, evidence has been accumulating that SLAM family members are potential targets for inflammatory and autoimmune diseases. This review will discuss the structure and function of the SLAM family receptors and SAP family adaptors, their role in immune regulation, and potential approaches to target this family of receptors therapeutically.

Introduction

The SLAM family of receptors consists of nine distinct members. These members include: SLAMF1 (SLAM or CD150), SLAMF2 (CD48), SLAMF3 (Ly-9 or CD229), SLAMF4 (2B4 or CD244), SLAMF5 (CD84), SLAMF6 (Ly108 in mice, NTB-A or SF2000 in humans), SLAMF7 (CRACC, CD319 or CS1), SLAMF8 (CD353 or BLAME), and SLAMF9 (SF2001 or CD84H). In terms of classification, SLAMF2, SLAMF8 and SLAMF9 are not considered full members of the SLAM family and can be designated as atypical (Table 1) (1). This is due to the fact that SLAMF2, SLAMF8, and SLAMF9 do not share homology in their cytoplasmic domains when compared to the rest of the typical SLAM family (Table 1). All the receptors in this family are assigned to the CD2 superfamily immunoglobulin (Ig) domain-containing molecules and are known to be widely expressed on hematopoietic cells, where most cells express between 3 to 5 individual SLAM family members (2). Interestingly, although SLAM family receptors are considered to be homophilic, it has been reported that they could also bind to several morbilliviruses, such as the measles (3).

Table 1.

Classification and nomenclature of the SLAM family members.

Typical family members	Atypical family members
SLAMF1 (CD150)	SLAMF2 (CD48)
SLAMF3 (Ly-9 or CD229)	SLAMF8 (CD353 or BLAME)
SLAMF4 (2B4 or CD244)	SLAMF9 (SF2001 orCD84H)
SLAMF5 (CD84)
SLAMF6 (Ly108, NTB-A or SF2000)
SLAMF7 (CRACC, CD319orCS1)

SLAMF1 will form homodimers when examined via surface plasmon resonance (SPR), in a “head to head” fashion (4), meaning that the binding occurs between the membrane distal Ig V-like domains. SLAMF2 does not self-associate, in contrast to SLAMF3 which exhibits a similar self-association mode when studied by introducing point mutations in the Ig V-like domains (5). When studied by crystallography, SLAMF4, SLAMF5, and SLAMF6 bind to their respective self-ligands by their Ig V-like domain as well (1, 6–8). SLAMF7 is also a self-ligand that binds via its Ig V-like domain as determined by antibody binding analyzed by flow cytometry (9). SLAMF8 and SLAMF9 have no known ligands (1).

The SLAM family receptors and SLAM family adaptors are rapidly becoming a target of interest in the understanding and treatment of diseases that range from XLP (10–14) to HIV (15), and several types of cancers (2, 16–17). Moreover, it is becoming increasingly clear that SLAM family receptors play an integral role in the pathogenesis of autoimmune disorders (2, 16–21). Therefore, it is imperative to develop a holistic understanding of the mechanisms by which these proteins function in health and disease.

The structure of the SLAM family receptors

All the SLAM family proteins are considered part of the CD2 superfamily. The typical SLAM family receptors are type I glycoproteins that contain an amino terminal Ig like variable (V) domain which has no canonical disulfide bonds (Figure 1). They also have a membrane proximal constant 2 (C2) domain that houses two disulfide bonds (22). Among the typical members, SLAMF3 has a repeated IgV like domain and a C2 domain and thus has four extra cellular domains (Figure 1) (23–24). This occurs in a simple repeat of the V-like and C2-like domains (25). The cytoplasmic tails of the typical SLAM family members contain multiple immunoreceptor tyrosine-based switch motifs (ITSMs) (Figure 1) (26–27). These ITSMs act as ligands for the SAP family adaptors SAP, EAT-2 and ERT (28). The name ITSM comes from the need to differentiate it from immunoreceptor tyrosine activation motifs (ITAMs), which recruit Syk kinases, and immunoreceptor tyrosine inhibitory motifs (ITIMs), which frequently recruit phosphatases (1, 28). Each SLAM family member contains at least a single ITSMs and the homology across the ITSMs in the cytoplasmic domain is TxYxxV/I/L/ (where x is any amino acid).

Figure 1. The structure of the typical SLAM family receptors.

The SLAM family receptors are type I glycoproteins that contain an amino terminal Ig like variable (IgV) domains, a membrane proximal constant 2 (C2) domains, and phosphorylated tyrosines (Y), some of which are part of immune tyrosine switch motifs (ITSM) sequences (TxYxxV/I/L).

In regards to the atypical SLAM family members, SLAMF2 differs its cytoplasmic region from the other members, although it is analogous in the ecto domain. This receptor is held to the cell surface by means of a glycosylphosphatidylinositol (GPI) anchor (29), has no cytoplasmic domain (30), and thus lacks any ITSMs. However, this receptor is the only of a ligand for SLAMF4 and a potent co-stimulating receptor that is expressed on nearly all hematopoietic cells (30–32). As mentioned, SLAMF8 and SLAMF9 have only a short cytoplasmic domain and lack ITSMs (33–34).

The SAP family adaptors bind to the SLAM family receptors

The SAP family adaptors are comprised almost exclusively of a Src homology 2 (SH2) domain. The SAP adaptor family includes SAP, EAT-2, and ERT (Table 2). All of the SAP family members are cytosolic proteins and are expressed exclusively in hematopoietic cells. SAP is expressed in human and mouse T cells, natural killer (NK) cells, NK T cells as well as platelets (Table 2) (14, 25–26, 35–39). SAP is lowly expressed in B cells, both in humans and mice (40). EAT-2 is also expressed in humans and mice and typically originates in NK cells (37, 41) but can also be found in dendritic cells (DCs) and macrophages (25) where it is lowly expressed (41). ERT is found only in mouse NK cells as it exists as a pseudo gene in humans (25). While SH2D1A SAP is located on chromosome X, SH2D1B (EAT-2) and SH2D1B2 (ERT) are on chromosome 1 (Table 2).

Table 2.

The SAP family adaptor expression and chromosomal location.

Family member	Expression	Chromosomal location
SAP	T, NK-T, NK, B cells and platelets	X (Humans and mice)
EAT-2	DC, NK, and macrophage	1 (Humans and mice)
ERT	NK	1 (Mice and pseudo gene in humans)

SAP tends to bind to ITSMs (homology to TxYxxV/I/L/T) (42), located in the cytosolic tail of the SLAM family receptors but will bind preferentially to the specific ITSM sequence of TIYxxV/I/L/T (43). EAT-2 is not known to have a preference for specific ITSM sequence (44). This is exemplified by the fact that EAT-2 will bind to SLAMF7, but SAP will not (17). SLAMF7 contains only one ITSM (Figure 1) with the sequence of TEYxxV/I/L/T. Moreover, EAT-2 will associate with SLAMF1, SLAMF3, SLAMF4, SLAMF5, and SLAMF6 (25), none of which contain a TEYxxV/I/L/T ITSM. It is not clear what are the binding preferences of ERT are.

In addition to the SAP family adaptors, several SH2 domain containing phosphatases can bind to the ITSMs of the SLAM family receptors. These include SH2 domain-containing protein tyrosine phosphatases SHP-1 and SHP-2, as well as the SH2 domain-containing inositol phosphatase SHIP-1 (45–46). It is widely accepted that these phosphatases are prevented from binding to ITSMs through competition with SAP on the same binding sites (25, 41). Current literature has extensively exhibited that SLAM family receptor functions in immunity are dependent on their interactions with SAP family adaptors as well as SH2 domain-containing protein tyrosine phosphatases (26, 37–38, 40, 45–48). The next sections will discuss both activating and inhibitory pathways that are differently regulated by these adaptors and phosphatases.

Activating pathways initiated by the recruitment of SAP adaptors to the SLAM family receptors

SLAM receptors can act as inhibitory receptors or as activating receptors, depending on the binding of SAP family adaptors to the ITSMs in the cytoplasmic domains of SLAM family receptors (49–50). This is speculated to occur in both NK-T cells and in follicular helper T cells (T_FH) (25). Fyn mediated phosphorylation of the ITSMs occurs after self-association of SLAM family members (50–54), and is required for SAP binding to most SLAM family members (50–51). At first it was not clear how the binding of SH2 domain that has no intrinsic function could alter the function of a receptor. However, this issue became elucidated when it was shown that SAP can compete with other SH2 domain containing proteins such as SHP-2 for binding to the same ITSMs on SLAMF4 (26). This means that one function of SAP is to act as a “cap” and to competitively prevent phosphatases from binding to ITSMs.

This will serve a dual function. First, it will act to block the binding of SH2 domain containing phosphatases such as SHP-1, SHP-2 and SHIP-1, which in some setting, act as inhibitory mediators. Second, SAP can enhance downstream signaling through recruiting of the Src family kinase Fyn (Figure 2). The binding of Fyn to SAP takes place via the arginine 78 residue (R78) of SAP binding to the SH3 domain of Fyn (55–57). The R78 residue is located outside the phosphotyrosine binding pocket, allowing for the simultaneous interaction between the SLAM ITSM and the SH3 domain of Fyn (57). This will trigger additional downstream signaling events leading to cellular activation.

Figure 2. Dual signaling functions of SLAMF6.

Under SAP recruitment to phosphorylated ITSMs, the receptor behaves as activating receptors (A) while the kinase Fyn will bind to SAP via SAP’s arginine 78 residue. Under SHP-1, SHP-2 and SHIP-1 recruitment, the receptor behaves as inhibitory receptors (B). SLAMF6 can also recruit EAT-2 (C) through a different tyrosines, further leading to cellular activation as a result of SHIP-1 phosphorylation. The inhibitory kinase Csk (D) can recruit additional phosphatases via its phosphorylation of additional ITSMs.

The signaling events following the formation of SLAM-SAP-Fyn complex will differ depending on the type of SLAM family receptor. For example, in the case of SLAMF1 (43, 53), that can bind to SAP at low affinity even without tyrosine phosphorylation, the SLAM-SAP-Fyn formation will lead to recruitment and phosphorylation of SHIP-1 docking protein 1 (Dok1), docking protein 2 (Dok2), and the exchange factor RasGAP (58). Furthermore, it has been reported that in CD4 T cells SLAMF1 engagement will lengthen the conscription of PKCθ to the site of contact between CD4 T cells and antigen presenting cells (APCs) (59). Also, NF-κB, a protein that controls transcription of DNA, cytokine production and cell survival, has enhanced activation during this process (59), leading to IL-4 secretion (1).

Moreover, preventing SAP association with other members such as SLAMF3, SLAMF4, and SLAMF5 act as inhibitory receptors (45), implying that at least in this context, the presence SAP adaptors control the cell activation states. Interestingly, the same SLAM-SAP-Fyn complex in the context of SLAMF6 (60) will lead to the phosphorylation of the small G protein Vav-1 (47). It has been further demonstrated that the removal of SAP from the same complex will alter the cell cytotoxic function, from targeting hematopoietic cells to targeting non-hematopoietic cells in a NKG2D dependent manner (47).

It seems as though the type of SLAM in question is the determining factor for which SH2 domain containing phosphate will bind to the ITSMs. For example, SLAMF4 has a propensity to bind to SHIP-1 when studied in NK cells (61) and SLAMF6 has a tendency to bind to SHP-1 in T cells (62). Furthermore, SAP deficiency has been shown to inhibit polarization of T_FH cells, B cells functions, and initiating germinal center (GC) formation (48) via the interaction of SAP with SLAMF5 (63–64). This can lead to the prevention of lupus like symptoms (3). Although the process for how exactly SAP deficiency gives rise to T_FH cell dysfunction is not yet fully understood, a study has shown that in the absence of SAP, Ly108 (the mouse analogue to SLAMF6) becomes an inhibitory receptor in T_FH cells (62). SAP expression is also required in T cells, and partly in NK cells or NK T cells, to elicit a protective immune response. This was shown when floxed mice that have had SAP specifically deleted from T cells exhibited the same defects in humoral immunity as SAP deleted mice (40). Furthermore, SAP deficient mice show a deficiency in TH2 cytokine production in CD4 T cells (65–66). Another interesting facet to SAP regulation is that it binds directly to CD3ζ via the first membrane proximal ITAM to regulate T cell receptor (TCR) signaling in activated T lymphocytes (67). This may play an important role in disease where SLAMF3 and SLAMF6 are over expressed in T lymphocytes (68) as the over expression of SAP ligands may alter SAP trafficking and cause T cells to become over activated. This implies that SAP plays a more ubiquitous role in T cell activation than previously thought, allowing for the association or non-association of other molecules, such as inhibitory factors to the ITSMs of the SLAM family receptors which will be discussed next.

The recruitment of inhibitory factors to SLAM family receptors

As mentioned above, the SLAM family members SLAMF4 and SLAMF6 are known to recruit the SH2 domain-containing phosphatases, SHP-1, SHP-2 and SHIP-1 as well as the inhibitory kinase Csk (Figure 2) (1). SAP is known to prevent these interactions from occurring (13, 26, 60) by binding to the same phosphorylated ITSMs. Moreover, the kinase Csk is known to recruit additional phosphatases via its phosphorylation of additional ITSMs of SLAMF4 (69). Csk also phosphorylates and inhibits other members of the Src family kinases, further supporting downstream inhibitory signals. One more important layer of complexity is that SAP function differs depending on what tyrosine is binding to. In SLAMF4 it is thought that SAP binding to tyrosine 309 will block the interaction with SHP-2 (70). Moreover, SLAMF2 on target cells has been shown to behave as an inhibitory receptor when binding to SLAMF4 on NK cells (71–73). This behavior has also been reported with SLAMF7 (54). Consequently, and at least in some setting, it is likely that the physiological function of some SLAM family members is inhibitory by nature.

It has been witnessed in NK cells that Vav-1 is the primary substrate of SHP-1 (47). Vav-1 is an important initiator of intracellular Ca²⁺ flux and cytokine secretion in T cells and NK cells (74–75). Therefore, SLAMF6 as well as other SLAM family receptors may play roles as a docking site for phosphatases that control activation in NK cells and T cells by inhibition of downstream regulators such as Vav-1, suggesting that SLAM receptors themselves do not control activation states, rather they provide a platform for activation regulation by functioning as adaptors.

Differential binding of EAT-2 and ERT to SLAM family receptors

EAT-2 shares around a 50% homology in amino acid sequence with SAP in both human and mice. However, these two proteins differ in several key ways. Unlike SAP, EAT-2 and ERT do not have the arginine residue at position 78 that is required for binding to the kinase Fyn and are not coupled to Vav-1 (2). Rather, EAT-2 and ERT both have additional tyrosines; mouse EAT-2 and ERT have two additional tyrosines and human EAT-2 has only 1 (ERT is not known to be expressed in humans) (37). Furthermore, these tyrosines are located in the carboxy-terminal tail and may participate in the activating or inhibitory signaling of EAT-2 and possibly ERT (37, 54). In studies involving NK cells from mice, it has been shown that EAT-2 acts as an inhibitor to NK cell activation by its association with SLAMF4 (69). Also, it has been shown in EAT-2 deficient mice, NK cells secrete more interferon gamma (IFN-γ) and IL-2 downstream of the receptors CD16, NKG2D, Ly49D and SLAMF4 (37). The same study shows a similar trend in ERT deficient mice, although the increase in killing was seen only toward hematopoietic cells (such as YBDD cells) (37). Conversely, EAT-2 has been shown to take part in the activation of human NK cells when it associates with SLAMF6 (76). This may be due to the fact that it can bind to SLAMF6 simultaneously with SAP (76); EAT-2 binds to the tyrosine 285 in SLAMF6. This perhaps allows it to block SHP-1, SHP-2 and SHIP-1 from binding (27). Moreover, the ITSMs that bind EAT-2 are different in SLAMF4 and SLAMF5 as well as the aforementioned SLAMF6 (27, 77–78). Altogether, the dichotomous effect of EAT-2 and ERT association with SLAM family members warrants further investigation (79–82).

The role of SLAM and SAP family members in human diseases

Conceivably the most discussed human disease in which SAP deficiency is the cause of is XLP. Patients with XLP exhibit an increase in immune response to Epstein Bar virus (EBV) infection (10, 12). This is characterized by an overstated and in many times fatal infectious mononucleosis-like disease secondary to polyclonal B and CD8 T cell proliferation with massive infiltration to the liver and to the bone marrow. This is due to the fact that XLP1 patients fail to clear infected and reactive B cell clones (1, 11). This inevitably leads to hepatic necrosis and bone marrow aplasia and some additional evidence of hemophogocytosis has been suggested (83). Furthermore, B and T cell dysfunction, unrelated to EBV infection, also occurs in XLP1 manifested by hypogammaglobulinemia (84–85). Research has come about to describe the mechanism that causes the improper interactions between CD8 T cells and B cells in XLP1 patients. This uncommon genetic disorder is often fatal as patients develop hemophagocytic lymphohistiocytosis (HLH) syndrome (11). HLH is marked by a hyper activation of immune cells that attack healthy tissue in the bone marrow, liver, spleen or in the lymph nodes (79). Patients with XLP experience swelling of lymph nodes, an enlarged liver and spleen, and hepatitis. Some patients with XLP develop recurrent B-cell non-Hodgkin’s lymphoma even in the absence of EBV infection (80). Technically speaking there are two types of XLP: XLP1 and XLP2. These two types of XLP share homology in the clinical symptoms but not in the genetic root cause. XLP1 is caused by a mutation in the SH2D1A gene (13, 14), the gene that encodes SAP, whereas XLP2 patients exhibit X-linked inhibitor of apoptosis (XIAP) deficiency which is caused by BIRC4 mutations (81–82).

It has been demonstrated in human XLP patients that blocking SLAM family interactions by antibodies restores T cell function against B cell targets that also express SLAM family members. Furthermore, a synergistic effect was witnessed when antibodies were used to block SLAMF4 and SLAMF6 interactions (19). This follows from the fact that SLAMF2, the ligand for SLAMF4, is upregulated on EBV infected B cells (31). Also, it is important to note that XLP1 patients exhibit defects also in the functions of NK-T and NK cells (41). Furthermore, EBV has been discovered to have involvement in rheumatoid synovitis (86–87). Moreover, SLAMF2 is highly up regulated in EBV transformed B cells; which will induce NK cell activation via interaction with SLAMF4 (88). However, rheumatoid arthritis (RA) patients exhibit a lack in SAP association to SLAM. Therefore, this may be the reason for the inability of T cells and NK cells to clear EBV-infected synovial cells and B cells as seen in patients with RA (87–88). A similar mechanism may be at play with regards to patients with XLP. Lastly, patients with RA are at a much higher risk of myocardial infarction (89). For this condition patients are prescribed TNF-α blockers. However, this has the unintended consequence of causing autoimmunity via the lowering of SAP (90). Given the relation that both RA and XLP have to SLAM and SAP this topic warrants further investigation.

Another autoimmune disease in which SLAM takes part in is systemic lupus erythematosus (SLE). The sleb1 locus corresponds to the SLAM genes (SLAMF1 through SLAM) and is located on chromosome 1 (1, 21, 33, 35). This locus takes part in SLE pathogenesis due to polymorphisms in Slamf6 as was shown in 129Sv mice when compared to C57BL/6 mice (21). This results in augmented signaling by the SLAMF6 receptor and changes in B and T cell functions that give rise to inflammatory symptoms (21, 91–94). Furthermore, it is now understood that both SLAMF3 and SLAMF6 are involved in SLE as shown in human T cells collected from lupus patients (68). This occurs via SLAM receptors co-stimulating TCR when both receptors are engaged with their respective ligands simultaneously, resulting in downstream enhanced IL-17 production (68). Furthermore, SLAM co-stimulation with TCR advances Th17 differentiation in naive CD4 T cells. This is a very import finding as patients who exhibit SLE tend to have higher expression levels of SLAMF3 and SLAMF6 on their T cells in correlation with disease severity (68). Evidence has also been elucidated that SAP is involved in SLE. This was shown in a MRL/lpr mouse model that exhibited a loss in the SAP gene leading to defects in auto antibody production (95–96). The contribution and the function of SLAMF6 in NK cells to lupus pathogenies needs further investigation (47).

Clinical advantage of targeting the SLAM family receptors

The kinase inhibitor sorafenib was given FDA approval to treat hepatocellular carcinoma (HCC) in 2007 and since then has shown to increase patient survival by 44% (97–100). It acts as a multi kinase inhibitor of several protein tyrosine kinases such as VEGFR, PDGFR and Raf-family kinases (97–99). Despite the relative successes of this drug there remain multiple patients whose resistance to sorafenib leads to therapeutic failure. Interestingly, SLAMF3 shows reduced expression in cancerous liver tissue and in vitro introduction of SLAMF3 inhibits tumor growth (18), suggesting that SLAMF3 expression might predict resistance to sorafenib. Moreover, it has been attempted to determine if SLAMF3 was inhibiting the ability of sorafenib to treat HCC patients. Future research will have to determine if there is any use to combine this drug with anti-SLAMF3 interventions.

SAP deficient NK cells extracted from mice have been demonstrated to target and kill non-hematopoietic cells (47). This targeting allows NK cells to lyses B16F melanoma cells in a manner that is SAP dependent (47). This study also suggested that the upstream binding partner SLAMF6 is a potential target in cancer immunotherapy, allowing NK cells to act on both hematopoietic and non-hematopoietic transformed cells (47). Moreover, SLAMF6 has promise also as a biomarker for HIV infection. HIV infected cells will exhibit a down regulation of the major histocompatibility complex class I molecules (MHC-I) (101) which typically allows NK cells to sense and kill such cells (102). However, HIV infected cells are not killed by NK cells despite this fact (103). Moreover, it has been shown that this is the case because HIV infected T cells will undergo an up regulation of human leukocyte antigen (HLA)-C and HLA-E, which bind to inhibitory NK cell receptors and inhibit NK cell activity (104). Previous studies have revealed that lymphoblastic leukemia cells are able to avoid NK cell detection because they down modulated the expression of SLAMF4 and SLAMF6 (105). Therefore, it was investigated whether or not SLAMF4 and SLAMF6 play a role in NK cell killing of HIV infected cells. When a ⁵¹Cr release assay was conducted on NK cell targets (autologous infected CD4 cells) with and without blocking antibodies for SLAMF4 and SLAMF6 on NK cells, it was demonstrated that NK cell killing of HIV infected cells depends on SLAMF4 and SLAMF6 interactions (15). Furthermore, it was demonstrated that SLAMF4 and SLAMF6 are down regulated during infection of the same cells with HIV (15).

Evidence has also begun to compile regarding targeting SLAMF7 for some types of cancer. This receptor is over expressed on multiple myeloma cells (93), a cancer that is formed by malignant plasma cells (94). In light of these findings the IgG1 non-blocking antibody elotuzumab was developed (16) to recognize specifically SLAMF7 (17). The antibody acts by engaging SLAMF7 on multiple myeloma cells via the Fab portion of the antibody. The Fc portion of elotuzumab will bind to the activating receptor CD16 on NK cells which increases their cytotoxicity towards multiple myeloma cells (2) and has recently been approved in combination with lenalidomide and dexamethasone to treat patients with relapsed or refractory multiple myeloma. SLAMF7 has also exhibited therapeutic potential for patients with RA. Targeting the antigen CD20 with anti-CD20 monoclonal antibodies, which depletes B cells from the circulation (106), has been shown to be insufficient in some patients. These patients have a persistence of CD20 negative plasmablast and plasma cell populations (107). However, these cells strongly express SLAMF7 (16). Therefore, a humanized antibody dubbed PDL241 was developed to target SLAMF7. PDL241 was subsequently shown to inhibit the production of immunoglobulins in a mode that was an Fc-dependent in vitro by killing plasmablasts and plasma cells, but not B cells, in peripheral blood mononuclear cells (PBMC) cultures (108). It was further shown in an in vivo rhesus monkey model that PDL214 treatment reduced the severity of joint-related disease parameters via the reduction of IgG and IgM antibodies (108).

Concluding remarks

Over the past decade it has become increasingly clear that the SLAM family of receptors are important regulators of immune response. This has been demonstrated on most types and subsets of lymphocytes, particularly in NK cells, although SLAM family receptors are also required for T and B cell immunity. At first it has been shown that the receptor functions of SLAM family members such as SLAMF4 are SAP dependent. Furthermore, the function of the receptor can change from inhibitory to activating depending on the context of the interactions and beyond. The context of the interactions can be thought of as function switch that is due to the binding of SH2 domain containing proteins such as SAP or SHP-1 phosphorylated tyrosines in the cytoplasmic domain of the SLAM family proteins. This may be further affected by the concentration of SAP family adaptors as well as the expression levels and sub cellular location of the various SLAM family receptors or the availability of other SAP binding proteins such as CD3ζ. This is further complicated when it is considered that SLAMF2 binding to SLAMF4 has been shown to act as an inhibitory mechanism for NK cells when SLAMF2 is expressed on NK cell targets. As a result, it could be that SAP recruitment to SLAMF4 after the interaction between SLAMF2 and SLAMF4 shut down NK cell activation. Alternatively, phosphatases such as SHP-1, SHP-2, and SHIP-1 may be primarily recruited after the interaction between SLAMF2 and SLAMF4. This potentiates SHP-1, SHP-2, and SHIP-1 as NK cell activators. However, SLAM family members tend to only interact with one another via their IgV like domains, with the exception of SLAMF4, which will interact with SLAMF2.

SLAM family receptors have been identified in many types of disease pathologies. Patients who exhibit XLP1 have a marked inability for their T cells to clear B cells (109). Also, the B cells of XLP1 patients demonstrate a distinct phenotype (CD10⁺CD24^HighCD38^HighCD5⁺Bcl2⁻) that makes them behave more as immature B cells (110). Therefore, these cells will exhibit a reduction in survival, proliferation, differentiation, chemotaxis, and antibody production compared with mature B cells (110). This will inevitably make them less able to clear infection. Moreover, both T cells and B cells express SLAMF6. When SLAMF6 interactions are blocked in XLP1 cells, normal T cell clearance of B cells can be seen. This is an example of SLAMF6 acting as a receptor that inhibits T cell function; although, B cells express little to no SAP, which can affect the role of SLAMF6 with respect to their cytotoxicity. However, in SAP deleted cells, SLAMF6 deletion will halt NK cell cytotoxicity against non-hematopoietic targets. These targets do not express ligands for SLAMF6. Therefore, SLAMF6 may be acting as an adaptor molecule by altering the trafficking of SHP-1. To further complicate this situation SLAMF4 exhibits similar behavior, but is not only a self-ligand; SLAMF4 is seen to affect B cell clearance in a synergistic manner with SLAMF6 in XLP1, but it does this via interaction with SLAMF2 although it can still interact in a homotypic manner.

To summarize, it is important to understand the nature of the contexts in which SLAM family members switch their function from inhibitory to activating receptors and also the ramifications of this functional alteration. Furthermore, investigation should be done to see if this switch of function is due to SLAM family members behaving as receptors or adaptors. A clear pattern of observations has elucidated the importance SLAM family proteins in immune responses. Now a clear understanding of these mechanisms must be understood in order to design interventional approaches to treat both cancer and other inflammatory conditions.

Table 3.

The SLAM family receptor ligands, expression and interaction with SAP family adaptors.

Family member	Ligand	Expression	SAP binding	EAT-2 binding	Homology to SLAM family
SLAMF1	SLAMF1, Measles virus	Activated T, B, DC, Macrophage, Platelet, GC TFH	+	+	Ecto/cytoplasmic domain
SLAMF2	SLAMF4	T, Activated T, B, DC	−	−	Ectodomain
SLAMF3	SLAMF3	Activated T, B, DC, Macrophage, Platelet	+	+	Ecto/cytoplasmic domain
SLAMF4	SLAMF2 (CD48)	CD8T, NK, DC, Macrophage	+	+	Ecto/cytoplasmic domain
SLAMF5	SLAMF5	T, TFH, B, NK, DC, Macrophage, Platelet	+	+	Ecto/cytoplasmic domain
SLAMF6	SLAMF6	T, TFH, B, NK, DC, Neutrophil	+	+	Ecto/cytoplasmic domain
SLAMF7	SLAMF7	Activated T, B, NK, DC, Macrophage	−	+	Ecto/cytoplasmic domain
SLAMF8	Unknown	Macrophage	−	−	Ectodomain
SLAMF9	Unknown	T, B, NK, DC	−	−	Ectodomain