Negative Regulation of Humoral Immunity Due to Interplay between the SLAMF1, SLAMF5, and SLAMF6 Receptors

Whereas the SLAMF-associated protein (SAP) is involved in differentiation of T follicular helper (Tfh) cells and antibody responses, the precise requirements of SLAMF receptors in humoral immune responses are incompletely understood. By analyzing mice with targeted disruptions of the Slamf1, Slamf5, and Slamf6 genes, we found that both T-dependent and T-independent antibody responses were twofold higher compared to those in single knockout mice. These data suggest a suppressive synergy of SLAMF1, SLAMF5, and SLAMF6 in humoral immunity, which contrasts the decreased antibody responses resulting from a defective GC reaction in the absence of the adapter SAP. In adoptive co-transfer assays, both [Slamf1 + 5 + 6]−/− B and T cells were capable of inducing enhanced antibody responses, but more pronounced enhancement was observed after adoptive transfer of [Slamf1 + 5 + 6]−/− B cells compared to that of [Slamf1 + 5 + 6]−/− T cells. In support of [Slamf1 + 5 + 6]−/− B cell intrinsic activity, [Slamf1 + 5 + 6]−/− mice also mounted significantly higher antibody responses to T-independent type 2 antigen. Furthermore, treatment of mice with anti-SLAMF6 monoclonal antibody results in severe inhibition of the development of Tfh cells and GC B cells, confirming a suppressive effect of SLAMF6. Taken together, these results establish SLAMF1, SLAMF5, and SLAMF6 as important negative regulators of humoral immune response, consistent with the notion that SLAM family receptors have dual functions in immune responses.


INTRODUCTION
The humoral immune response is crucial for protecting individuals from many infections and eliminating foreign substances. Antibody responses can be induced in a T cell dependent or T cell independent manner. In T-independent immunity, the antibody response occurs directly after B cell activation in T cell deficient mice. In contrast with T-independent responses, T-dependent responses require T cell help for B cell activation and maturation. In the presence of T cell help, B cells undergo robust proliferation and somatic hypermutation of their variable region genes and differentiate into high affinity memory B cells and long-lived plasma cells in the germinal centers (1,2). Although several subsets of CD4 + T helper cells may be implicated in T-dependent humoral responses, it becomes clear that the follicular helper CD4 + T cell subset [T follicular helper (Tfh)] is a major B cell help provider (3)(4)(5)(6). Tfh cells exhibit a phenotype distinct from that of other effector CD4 + T helper cells, as they express the transcription factor B cell lymphoma 6 (Bcl6) that is necessary for the development of Tfh cells and inhibits expression of genes critical for development of other T helper cells (4,(7)(8)(9). Although Bcl6 expression by pre-Tfh cells is required, it is not sufficient in vivo for full polarization of Tfh cells. In fact, multiple molecules have been shown to be involved in the differentiation of Tfh cells (3,4,6). In addition, Tfh development is highly dependent on B cell responses, as Tfh cells are not found in B cell deficient mice (7,10,11). These findings indicate that, through their interaction, GC B cells and Tfh cells reciprocally provide each other with signaling for survival, proliferation, and differentiation.
The signaling lymphocytic activation molecule family (SLAMF) includes nine structurally related Ig-like proteins that are differentially expressed on the surface of hematopoietic cells (12). SLAMF receptors have been shown to function as co-stimulatory molecules and to modulate the activation and differentiation of a wide array of immune cell types involved in both innate and adaptive immune responses (12)(13)(14). While most SLAMF receptors serve as self-ligands, SLAMF2 and SLAMF4 interact with each other. Six SLAMF receptors (SLAMF1, SLAMF3, SLAMF4, SLAMF5, SLAMF6, and SLAMF7) carry one or more copies of an immunoreceptor tyrosine-based switch motif (ITSM) in their cytoplasmic tails. This signaling switch motif can recruit SH2 domain-containing signaling molecules such as SLAM-associated protein (SAP) (15). SAP is a cytoplasmic adapter molecule with a single Src homology 2 domain and a small carboxy-terminal region. The SAP family consists of three members: SAP expressing T, NK, and NKT cells, and EAT-2A and EAT-2B (murine) expressing NK cells and APC (12,16). There is accumulating evidence that SAP and EAT-2 can function as signaling adaptors that link SLAMF receptors to active signaling molecules such as the Src www.frontiersin.org family protein tyrosine kinases Fyn and PI3K (15,(17)(18)(19)(20)(21). SAP and EAT-2 have also been shown to act as blockers to outcompete SH2 domain-containing inhibitory molecules SHP1, SHP2, and SHIP1 (22)(23)(24)(25)(26)(27)(28).
Deficiencies in the gene that encodes SAP (SH2D1A) result in a primary immunodeficiency called X-linked lymphoproliferative disease (XLP) (29)(30)(31). Patients with XLP suffer from fatal infectious mononucleosis, malignant B cell lymphomas, and dysgammaglobulinemia. Defects in humoral responses and lack of germinal center formation are observed in XLP patients and in virally infected or immunized SAP-deficient mice (32)(33)(34)(35)(36). Considerable evidence indicates that the humoral immune response defect in XLP patients and SAP −/− mice stems from a defect in CD4 + helper cells because T-dependent antigen responses are defective and are restored after reconstitution with WT CD4 + T cells, but not WT B cells (32,34,35). However, the role of SAP in T-dependent humoral responses remains unclear. The SLAM/SAP/FynT axis regulates IL-4 producing Th2 differentiation, as demonstrated by the observation that Slamf1 −/− and SAP −/− mice have defective Th2 cytokine secretions (37)(38)(39). IL-4 is known to stimulate B cell antibody responses and Ig class switching, but the R78A mutant SAP mice can mount normal T-dependent antibody responses even though this mutant SAP molecule loses Fyn binding motif R78 (40). Recently, Qi and coworkers elegantly showed that SAP-deficient CD4 + T cells cannot form lasting mobile conjugate pairs with cognate B cells in the germinal center while the interaction between SAP-deficient T cells and DC is not affected (41). Since a sustained T-B conjugate allows their comprehensive activation and subsequent differentiation to Tfh cells and GC B cells, unstable T-B cell conjugates may contribute to humoral immune deficiency in SAP-deficient mice and XLP patients.
Compared to severe immunodeficiencies in SAP −/− mice, single ablation of SLAMF receptors causes a mild phenotype (40,(42)(43)(44). When mice deficient in SLAMF1 (the prototypic member of the SLAMF receptors) were infected with LCMV, Tfh cell differentiation, germinal center development, and the acute or long-term anti-viral antibody responses were comparable between LCMV-infected WT and Slamf1 −/− mice (40). Similar to Slamf1 −/− mice, Slamf3 −/− , Slamf5 −/− , and Slamf6 −/− mice showed no defects in response to LCMV (28,42). This suggests that functional redundancy exists among the SLAMF receptors, which has been confirmed in NKT cell development in pseudo Slamf [1 + 6]-deficient bone marrow reconstituting mice (45) and recent Slamf[1 + 6] −/− mice (46). For more than a decade, investigating the role of SLAMF receptors in SAP-mediated signaling has been difficult due to an inability to generate double or multiple SLAMF receptor deficient mice, as the receptors are closely located on the same chromosome (12). To define roles of SLAMF receptors as well as how they interact with one another in humoral immune responses, we generated Slamf [1 + 6] double knockout and Slamf[1 + 5 + 6] triple knockout mice using a two-time gene targeting technique and Cre/LoxP system. Surprisingly, we found that the combined absence of SLAMF1, SLAMF5, and SLAMF6 results in higher antibody production in response to both T-dependent and T-independent antigens. In addition, the administration of anti-SLAMF6 monoclonal antibody also impairs humoral immune responses in vivo. These observations suggest that SLAMF1, SLAMF5, and SLAMF6 function as negative regulators in T-dependent and T-independent antibody responses.

MICE
To generate Slamf [1 + 6] −/− and Slamf[1 + 5 + 6] −/− mice, a B6 bacterial artificial chromosome clone (B6 BAC clone #RP23-77A8) containing the Slamf1 and Slamf6 genes was used to construct a targeting vector with a neomycin resistant cassette flanked by two LoxP sites. SLAMF6 ES cell clones heterozygous for the mutation were generated by standard methods. To generate Slamf1 and Slamf6 double-deficient mice, we used a SLAMF1 targeting vector to retarget the previously generated SLAMF6 mutant ES cell clone that was known to give germline transmission with extremely high frequency. Co-integration of the two targeting vectors on the same chromosome was assessed by in vitro transfection-targeted ES cell clones with a Cre recombinase expression vector. Deletion of the whole Slamf1, Slamf5, and Slamf6 locus was confirmed by PCR (Figures 1A,B). B6 background Slamf5 −/− mice have been reported previously (46). Wild-type C57BL/6 (B6) mice were obtained from the Jackson Laboratory. Animal studies were conducted in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals and were approved by the Beth Israel Deaconess Medical Center Institutional Animal Care and Use Committee.

ANTI-SLAMF RECEPTOR ANTIBODIES
Rat anti-mouse SLAMF1 mAb (9D1) is specific for the extracellular region of mouse SLAMF1 (25). The anti-mouse SLAMF5 mAb was generated by fusing NS1 myeloma cells with spleen cells obtained from Armenian hamsters, which were immunized three times with mouse SLAMF5-Fc fusion protein (47). Mouse antimouse SLAMF6 mAbs (13G3 and 330) are specific for the extracellular region of mouse SLAMF6. Anti-SLAMF1, anti-SLAMF5, and anti-SLAMF6 mAbs used in our studies were purified by affinity chromatography (Harlan Bioproducts for Science). Anti-SLAMF6 F(ab') 2 fragments were generated from whole SLAMF6 mAb, using the Thermo Scientific Pierce F(ab') 2 Preparation Kit (PI-44988) according to manufacturer instructions.

IMMUNIZATION
Mice were immunized intraperitoneally (i.p.) with 40 µg of NPovalbumin (NP-OVA, Biosearch Technology) precipitated with Complete Freund's Adjuvant (Difco) or Alum. For eliciting TI antigen responses, mice were i.p. immunized with 20 µg of NP28-Ficoll (Biosearch Technologies) or with 10 µg NP-LPS. Mice were bled on day 9 (for T-dependent antigen) or day 7 (for Tindependent antigen) post-immunization. NP-specific IgG and IgM titers were determined by ELISA after serial dilutions of the serum.

ENHANCED T CELL DEPENDENT ANTIBODY PRODUCTION IN
Although most SLAMF receptors are expressed on the surface of T and B cells, ablation of single SLAMF genes does not lead to significant defects in germinal center formation after protein immunization or viral infection (27,42,44). In contrast, the absence of SAP, the SLAMF specific adaptor, leads to a severe defect in humoral response (14,32,34), which suggests functional redundancies in the control of antibody responses by SLAMF receptors. To test this hypothesis, we compared NP-specific antibody pro- Slamf [1 + 6] −/− mice, which had been immunized with NP-OVA in CFA, produced similar amounts of NP-specific serum IgM as WT mice (data not shown). However, the level of anti-NP IgG in the serum of Slamf[1 + 6] −/− mice was consistently higher, although statistical analysis did not reach significance ( Figure 3A). Surprisingly, the additional disruption of the Slamf5 gene significantly augmented the level of anti-NP IgG in Slamf[1 + 5 + 6] −/− mice ( Figure 3A) even though NP-specific IgM production was not altered (data not shown). Affinity maturation of NP-specific IgG was comparable between the mutant and WT mice (data not shown). Because Slamf[1 + 5 + 6] −/− mice produced higher NP-specific IgG compared to Slamf[1 + 6] −/− mice, we reasoned that SLAMF5 signaling might suppress antibody responses. To test this, we then immunized Slamf5 −/− mice with NP-OVA. In contrast to a previous report (41), Slamf5 deficiency had no effect on NP-specific antibody production or the development of Tfh cells or GC B cells (Figures 3B-F). Taken together, the data support the notion that SLAMF1, SLAMF5, and SLAMF6 cooperate in the negative regulation of T-dependent antibody responses.

THE COMBINED ABSENCE OF SLAMF1, SLAMF5, AND SLAMF6 ENHANCES ANTIGEN SPECIFIC PLASMA CELL EXPANSION, BUT HAS NO EFFECT ON THE DEVELOPMENT OF GC B CELLS, TFH CELLS, OR T FOLLICULAR REGULATORY (TFR) CELLS
As strong humoral immune responses, characterized by GC formation and long-lived plasma and memory B cells, are dependent on help provided by CD4 + Tfh cells (4,5,48), we next examined whether enhanced T-cell dependent antibody responses in  Given the increased antibody responses observed in Slamf[1 + 5 + 6] −/− mice, we hypothesized that the absence of SLAMF, SLAMF5, and SLAMF6 may enhance either plasma cell differentiation or their capacity to produce antibody. To test this hypothesis, we evaluated NP-specific plasma cells in NP-OVA immunized mice. Consistent with high antibody responses, the frequency of plasma cells (B220 + IgD -CD138 + ) was significantly increased in Slamf[1 + 5 + 6] −/− mice (Figures 4A,B).
Furthermore, flow cytometric analysis confirmed an increase in NP-specific plasma cells in immunized Slamf[1 + 5 + 6] −/− mice as compared to wild-type mice (Figures 4C,D). Together, the data indicate that the absence of SLAMF1, SLAMF5, and SLAMF6 has no effect on Tfh and GC B cell development, but that it appears to regulate development of antigen specific plasma cells.
A new Treg cell subset termed Tfr cells has recently been identified (49,50). These cells, which express not only CXCR5 and PD-1 but also the transcription factors Bcl6 and FoxP3, suppress both Tfh cells and GC B cells. As Tfr cell differentiation requires SAP expression (49), it is possible that the absence of SLAMF1, SLAMF5, and SLAMF6 might also cause a defect in Tfr cell development, thereby contributing to the enhanced antibody responses  (Figures 2A,B) that are known to participate in responses to T-independent antigens, we questioned whether T-independent antibody responses are affected in Slamf[1 + 5 + 6] −/− mice. Consequently, we examined the SLAMF mutant mice in response to NP-Ficoll, a classical synthetic TI-2 antigen that induces murine Ag-specific B cells to expand, differentiate, and produce NPspecific antibodies. As anticipated, the serum concentrations of NP-specific IgM and IgG3 in Slamf[1 + 5 + 6] −/− mice were increased on day 7 after immunization compared to those in WT mice (Figures 6A,B). In contrast, NP-specific IgM and IgG3 titers were comparable in WT and Slamf[1 + 6] −/− mice (Figures 6A,B). As responses to NP-LPS were not altered (data not shown), the enhanced TI-2 responses are specifically linked to the combined absence of SLAMF1, SLAMF5, and SLAMF6.

ANTI-SLAMF6 ANTIBODY INHIBITS HUMORAL IMMUNE RESPONSES
The observation that the ablation of SLAMF1, SLAMF5, and SLAMF6 enhances T-dependent and T-independent antibody responses suggests that they function as potential inhibitory molecules in humoral immune responses. In addition, a recent report indicates that SLAMF6 transmits inhibitory signaling in Tfh differentiation and NKT development in the context of the absence of SAP (27). To directly evaluate whether triggering of a single SLAMF receptor would initiate inhibitory signaling in humoral responses, we first sought to trigger SLAMF6 by using anti-SLAMF6 mAb in NP-OVA-immunized mice. As shown in Figure 7, the treatment of WT mice with anti-SLAMF6 (330) dramatically impaired NP-specific IgG production 9 days post-immunization of NP-OVA ( Figure 7A). As negative control, anti-SLAMF6 had no detectable effects on antibody production in Slamf[1 + 6] −/− mice ( Figure S4A in Supplementary Material). Noticeably, anti-SLAMF6 injected mice had a significant reduction in IgG high affinity antibody ( Figure 7B). Furthermore, NP-specific IgM production and its affinity maturation also were impaired in anti-SLAMF6 injected mice (Figures 7C,D). In correlation with the impaired antibody response, the frequencies and absolute numbers of splenic GC B cells (GL7 + Fas + ) (Figures 8A-C) and Tfh T cells (CXCR5 + PD-1 + ) (Figures 9A-C) were significantly reduced in anti-SLAMF6 injected mice. A reduced number of plasma cells was also observed in anti-SLAMF6 injected mice (Figures 8D,E), but effector B cell and T cell populations were equivalent in anti-SLAMF6 injected and non-injected mice ( Figure 8F and Figures 9D,E). In order to confirm the inhibitory effects of anti-SLAMF6 (330), another anti-SLAMF6 mAb (13G3) was also tested. As expected, a similar inhibitory effect was seen in 13G3-injected mice ( Figures S4A,B in Supplementary Material).
To address whether the administration of anti-SLAMF6 affects early commitment to antigen specific Tfh cells and GC B cells or late stages of humoral responses, immunized mice were injected with anti-SLAMF6 four days post-immunization, at which point T and B cells are already committed to becoming Tfh and GC B cells Frontiers in Immunology | T Cell Biology  (51). Interestingly, the late injection of mice with anti-SLAMF6 did not significantly reduce GC response and antibody production (data not shown). This suggests that the signal initiated by SLAMF6 has efficient inhibition in early Tfh and GC B cell differentiations, but has little effect on late Tfh and GC B cell expansion and antibody production. To evaluate whether the Fc portion of anti-SLAMF6 influences immune function, anti-SLAMF6 F(ab') 2 fragments were injected into mice along with NP-OVA immunization. Similar to intact anti-SLAMF6, anti-SLAMF6 F(ab') 2 caused a significant decrease in the percentage and number of GC B cells (Figures 10B,C) and Tfh cells (Figures 10D,E). In contrast, an impaired NP-specific antibody production was not observed in the anti-SLAMF6 F(ab') 2 injected mice. To exclude the possibility that NK cells mediate natural cytotoxicity against a variety of immune cells, we examined splenocyte phenotype and levels of www.frontiersin.org NP-specific antibody in NK-depleted WT mice following NP-OVA immunization and anti-SLAMF6 injection. Notably, the depletion of NK cells did not impact the capacity of anti-SLAMF6 to suppress antibody production or the development of Tfh cells and GC B cells (data not shown). Thus, severely impaired antibody production by the injection of anti-SLAMF6 is not due to NK cell-mediated ADCC.
Based on the studies using anti-SLAMF6, we next examined the ability of anti-SLAMF1 mAb and anti-SLAMF5 mAb to suppress humoral immune responses in NP-OVA immunized mice. We found that NP-specific antibody production was moderately, but consistently, reduced in anti-SLAMF1 injected mice ( Figure 7E). However, anti-SLAMF5 did not suppress antibody production ( Figure 7F). In addition, the frequencies of GC B cells and Tfh cells were comparable among the non-injected, anti-SLAMF1, and anti-SLAMF5 injected mice (data not shown). Thus, the results suggest that neither anti-SLAMF1 nor anti-SLAMF5 is alone sufficient to have a significant impact on humoral responses.

DISCUSSION
An important role of SAP in T cell help to B cells has been highlighted in defects in follicular T helper cell differentiation and lack of germinal center development in XLP patients and in virally infected or immunized SAP −/− mice (4,12,13,32,48,52). Although SAP has been proven to bind to ITSMs in most SLAMF receptors upon ligand stimulation, how these receptors function in the presence and absence of SAP is poorly understood, particularly in B cells. In this study, we provide in vivo evidence that SLAMF1, SLAMF5, and SLAMF6 synergistically suppress T-dependent and T-independent antibody responses, as Slamf[1 + 5 + 6] −/− mice, but not Slamf[1 + 6] −/− mice or single SLAMF mutant mice, exhibit a significant enhancement of antibody production. Strikingly, an adoptive transfer assay shows that the enhanced antibody responses depend largely on Slamf[1 + 5 + 6] −/− B cells, which indicates that the absence of SLAMF1, SLAMF5, and SLAMF6 induces intrinsic B cell activity. Furthermore, injection of mice with anti-SLAMF6 mAb dramatically reduced antibody responses accompanied by impairing Tfh cell and GC B cell development in spite of a less suppressive effect of anti-SLAMF1 mAb on antibody responses. Our results therefore point to a new and important mechanism by which SLAMF1, SLAMF5, and SLAMF6 regulate humoral responses in B cells. Maintaining sustained CD4 + T cell adhesion to B cells is required for Tfh differentiation and germinal center development, which allows for important signal transfer between T and B cells. SAP-deficient T cells fail to form a stable T-B cell conjugate (41) and thereby severely impaired development of Tfh cells and GC B cells becomes a hallmark in XLP patients and SAP −/− mice (14,32,34,41). Although most SAP-binding SLAMF receptors are highly expressed on resting and activated T and B cells and implicate a diverse array of lymphocyte functions, including sustained T-B cell conjugates (43,53), the deficiencies in single SLAMF receptors actually exhibit mild phenotypes. However, in contrast to our current study, Slamf5 −/− mice have previously been shown to have a defect in germinal center development and T-dependent antibody production in protein immunization (43). The reason that the www.frontiersin.org same B6 background Slamf5 −/− mice strains have different phenotypes is not clear. One explanation for this discrepancy is that the presence of selection marker (Neo) could influence neighboring genes in targeted loci (54). Alternatively, it remains possible that the discordant findings could reflect the different environmental conditions of animal facilities.
Because the functional redundancies in SLAMF-mediated signaling have been demonstrated in NKT cell development (45,46), we speculate that multiple deficiencies in SLAMF receptors may have a strong influence on humoral immune responses that are able to recapitulate most of the phenotypic alterations observed in SAP −/− mice. Surprisingly, in our in vivo studies comparing T-dependent antibody responses, loss of expression of SLAMF1, SLAMF5, and SLAMF6 receptors actually removed inhibitory signaling and resulted in higher antibody responses (Figure 3A). When the antibody responses in mice lacking Slamf [1 + 6] were compared to those in WT mice, there was also a consistent increase in NP-specific antibody titer, but the effect was less pronounced than that observed in the combined ablation of SLAMF1, SLAMF5, and SLAMF6. Although some variability in antibody responses between mutant and WT mice occurs, statistical analysis always reached significance. These findings indicate that the homophilic binding of SLAMF1, SLAMF5, and SLAMF6 synergistically transmits inhibitory signaling during humoral immune responses. The dual function of SLAMF receptors was initially reported in NK cell studies. In human NK cells, SLAMF4 predominantly functions as an activating receptor because engagement of SLAMF4 with SLAMF2 mediates NK cell cytotoxicity, cell proliferation, and cytokine secretion. However, in NK cells from XLP patients, the SLAMF2-SLAMF4 interaction fails to activate NK cells, but rather inhibits NK-medicated cytolysis (23). Besides SLAMF4, other SLAMF receptors such as SLAMF3, SLAMF5, and SLAMF6 become inhibitory molecules instead of activating receptors in mouse NK cells lacking SAP, EAT-2A, and EAT-2B (55). Consistent with these findings, a similar inhibitory effect of SLAMF6 on humoral responses and NKT cell development was reported in the context of the absence of SAP (27). Collectively, these observations suggest that most SLAMF receptors can mediate either positive or negative signaling, depending on the expression of SLAMF adaptors, SAP and EAT-2. Given that SAP and EAT-2A/B are not expressed in B cells, the homophilic interactions of SLAMF1, SLAMF5, and SLAMF6 between T and B cells or B and B cells would result in preferential binding to inhibitory signaling molecules (e.g., SHP1) in B cells because of the lack of competition of SAP and EAT-2 for ITSMs of SLAMF receptors. This idea is supported by our adoptive transfer assays, in which B cells from Slamf[1 + 5 + 6] −/− mice led to a further enhancement in antibody responses compared to the transfer of Slamf[1 + 5 + 6] −/− T cells. Furthermore, Slamf[1 + 5 + 6] −/− mice display higher antibody responses in the absence of T cell help when immunized with T-independent antigen NP-Ficoll, which directly indicates that an intrinsic B cell hyperactivation exists in Slamf[1 + 5 + 6] −/− mice.
Although the higher antigen specific antibody production does not accompany enhanced Tfh and GC B cell responses in the absence of SLAMF1, SLAMF5, and SLAMF6, increased development of plasma cells is consistently observed in NP-OVA immunized Slamf[1 + 5 + 6] −/− mice (Figures 4A-D). The mechanism regulating plasma cell differentiation is only partly understood. Two transcription factors Bcl-6 and Blimp-1 reciprocally modulate differentiation of GC B cells and plasma cells (52). Cytokines and chemokines also provide crucial survival signals to plasma cells (56). Far less is known about SLAMF receptor-mediated signals for plasma cell differentiation and function. However, in our experiments, we provide interesting evidence that SLAMF1, SLAMF5, and SLAMF6 negatively regulate either plasma cell differentiation and/or expansion in humoral immune responses. Further work will assess how SLAMF receptors are involved in plasma cell development at a molecular level.
The interesting finding in this study is that Slamf[1 + 5 + 6] −/− mice exhibit an increased frequency of MZ B cells. A cell-fate decision between follicular B cells and MZ B cells occurs in the transitional (T2) B cell stage, when T2 B cells differentiate into follicular B cells or MZ B cells after integration of BCR signal strength and signaling via other essential signaling molecules (57)(58)(59). Like in GC B cells, SLAMF1, SLAMF5, and SLAMF6 are highly expressed in transitional B cells and MZ B cells (ImmGen.org), and therefore, signaling resulting from their homophilic interaction may implicate differentiation, migration, or survival of MZ B cells. Since differentiation of transitional B cells was not altered in Slamf[1 + 5 + 6] −/− mice, signaling initiated by SLAMF1, SLAMF5, and SLAMF6 seems to play a critical role in controlling the development and/or survival of MZ B cells. It has been described that MZ B cells and B1 cells are prime B cell subpopulations responding to T-independent antigens (60,61). Interestingly, in spite of increased pools of MZ B cells, Slamf[1 + 5 + 6] −/− mice showed enhanced immune responses to the TI-2 Ag NP-Ficoll, but not to the TI-1 Ag NP-LPS. The fact that enhanced TI-antigen response is only limited to TI-2 antigens suggests that Slamf[1 + 5 + 6] −/− MZ B cell intrinsic activity, not number, seems to be more critical in determining the extent of humoral immunity. Thus, these results provide evidence that synergistic activity of SLAMF1, SLAMF5, and SLAMF6 may be implicated in functional activity of MZ B cells.
Complementary approaches with SLAMF receptor-deficient mice and SLAMF-specific antibodies are important for understanding the functions of their immunoregulatory pathways. Recently, SLAMF6 was found not only to constitutively associate with SHP1 in SAP-sufficient cells, but also to co-distribute with the CD3 complex. The ligation of SLAMF6 can reduce CD3ζ phosphorylation (53). Based on these findings, we hypothesize that crosslinking of SLAMF6 by anti-SLAMF6 mAb may cause high phosphorylation in its ITSM. Subsequently, protein tyrosine phosphatases and lipid phosphatases are preferentially recruited to SLAMF6, particularly in SAP and EAT-2 negative B cells. Surprisingly, we found that treatment of WT mice with anti-SLAMF6 almost recapitulates the phenotype observed in SAP −/− mice, which showed a marked defect in Tfh cell and GC B cell formation and reduced antibody production and affinity maturation. Interestingly, a significant defect in humoral response was not observed when mice were treated with anti-SLAMF6 four days after antigen exposure. This further indicates that SLAMF6mediated inhibitory signals have distinct roles in the early differentiation of Tfh cells and GC B cells. However, the administration of anti-SLAMF6 F(ab') 2 fragments did not fully suppress NP-specific Frontiers in Immunology | T Cell Biology FIGURE 11 | A model of SLAMF receptors: SHP1/2 action on the BCR during B cell activation. When B cells are activated, the ITSMs of SLAMF1, SLAMF5, and SLAMF6 recruit SHP1/2 and translocate these phosphatases to the vicinity of the B cell antigen receptor. Signaling from the BCR is thus down-regulated, maintaining proper response to antigens in humoral responses. When SLAMF1, SLAMF5, and SLAMF6 are deleted from B cells, inhibitory signaling mediated by SLAMF and SHP1/2 is dampened, which induces enhanced humoral responses.
antibody production even though the development of Tfh cells and GC B cells was significantly impaired (Figure 10). This difference may be due to a shorter half-life of anti-SLAMF6 F(ab') 2 , which could prevent it from maintaining sustainable triggering of SLAMF6 receptors during in vivo immune responses. Alternatively, the suppressive effect of anti-SLAMF6 mAb on humoral response might depend on the ability of anti-SLAMF6 Fc to bind to other accessory cells and crosslink the SLAMF6 receptors on Tfh cells and GC B cells. Such crosslinking is necessary for many surface molecules to initiate signaling events. Clearly, further studies are required to determine the contribution of other mechanisms such as complement and non-NK cell-mediated cytotoxicity in suppressing the activity of anti-SLAMF6 mAb. Compared to anti-SLAMF6 mAb, anti-SLAMF1 mAb has a milder, yet consistent, negative effect on antibody production, but anti-SLAMF5 mAb does not show any impact on humoral response. It seems contradictory to the functional redundancy we observed in Slamf[1 + 5 + 6] −/− mice. In fact, SLAMF6 has been shown to initiate dominant signaling in SAP −/− mice because ablation of the Slamf6 gene, but not the Slamf1 gene or Slamf5 gene, can rescue Tfh cell differentiation and antibody responses in SAP −/− mice (27,44). This indicates that SLAMF6 can complement deficiency of either SLAMF1 or SLAMF5 to facilitate inhibitory signaling in the absence of SAP. However, if Slamf1 and Slamf5 double mutations can be introduced into SAP −/− mice, impaired humoral responses may be partially restored. Similarly, the injection of mice with both anti-SLAMF1 mAb and anti-SLAMF5 mAb may also cause some reduction in antibody production.
By comparing T-dependent and T-independent antigen responses in Slamf5 −/− , Slamf[1 + 6] −/− , and Slamf[1 + 5 + 6] −/− mice, we demonstrated for first time that SLAMF1, SLAMF5, and SLAMF6 synergistically regulate humoral immune responses. Genetic interruption of SLAMF1, SLAMF5, and SLAMF6 results in enhanced antibody responses to T-dependent and T-independent antigens. In complementary studies, the administration of anti-SLAMF6 mAb further implicates SLAMF6 as a primary inhibitory member of the SLAMF receptors in antibody responses. The studies suggest that the ligation of SLAMF receptors in SAP-negative B cells (Figure 11) may preferentially recruit inhibitory signaling molecules to immunological synapses and control B cell responses during cognate interaction between T and B cells.