Previously, we showed that 2B4 is a dominant inhibitory receptor in SHIP-deficient NK cells that prevents efficient cytolysis of complex targets. We show in this study that 2B4 deficiency restores homeostatic control and cytolytic function to SHIP-deficient NK cells. However, 2B4−/−SHIP−/− NK cells still exhibit a profound disruption of their NK receptor repertoire and are compromised for induction of IFN-γ by several NK-activating receptors, including NKp46, NK.1.1, and NKG2D. In addition, we find that 2B4−/− NK cells have an extensively disrupted repertoire, including a supernormal frequency of NKp46+ NK cells. Consequently IFN-γ is induced on a much higher percentage of 2B4−/− NK cells following engagement of NKp46. We also find that both SHIP and 2B4 are required to prevent expression of Ly49B, a myeloid lineage MHC class I receptor not normally expressed by the NK lineage. Finally, when SHIP-deficient NK cells are on an H-2d background, they exhibit supernormal levels of Ly49A and possess normal cytolytic function against MHC-matched tumor targets and enhanced cytolysis of MHC mismatched tumor targets. However, despite normal or elevated cytolytic function, H2d SHIP−/− NK cells exhibit poor induction of IFN-γ like their H2b+ or 2B4−/− counterparts, demonstrating a uniform requirement for SHIP in induction of IFN-γ downstream of key NK activating receptors. These findings reveal a complex interplay of SHIP, 2B4, and MHC in the regulation of homeostasis, effector function, and repertoire formation in the NK cell lineage.

Natural killer cells express invariant receptors that are broadly categorized as activating or inhibitory. These receptors enable responses to viral infections, tumor cells, allogeneic grafts and damaged cells (1, 2). A fine balance, or lack thereof, between stimulatory and inhibitory signals emanating from these receptors dictates the function of an NK cell. Thus, a diverse and balanced NK receptor repertoire (NKRR) is extremely important in order for this lineage to respond to various immunologic challenges and to do so in a normal, effective manner. Aberrations in the expression of NK receptors (NKRs) or downstream signaling can lead to severe immune deficiency, as observed in SHIP-deficient mice (35). Upon engagement of an NK-activating receptor with its ligand, or cross-linking the receptor with an Ab, a functionally competent NK cell will produce cytolytic mediators (e.g., perforin), granzymes, and/or cytokines such as IFN-γ. NKRs are expressed in a variegated but overlapping fashion such that different cell subsets in the NK compartment elaborate different combinations of activating and inhibitory NK receptors. Varying the array of NKRs used by each subset increases the potential specificities of the NK compartment, while retaining tolerance to self (6). Therefore, it is essential to understand how this repertoire diversity is created, maintained, and remodeled (6). The acquisition of NKR expression is considered a semistochastic process, with the representation of an individual NKR largely determined by the relative strength of its promoter (7). This probabilistic model for NKR expression is based on inherent promoter features (7), as further supported by the demonstration that a genomic Ly49A transgene recapitulates the representation of Ly49A in the NK compartment (8). However, there is a wealth of evidence that the NKRR is not solely determined by relative promoter strength, because the presence or absence of NKR ligands, such as MHC class I (MHC-I), and mutations in intracellular signaling molecules can significantly impact the representation and expression of NKR (35, 9, 10). Thus, differential effects on the proliferation and/or turnover of NK subsets provide a secondary layer of regulation that determines the final composition of the NKRR (3, 6). Defining the ligands, receptors, and signaling molecules that constitute these secondary regulation pathways is required to fully understand how the NKRR is formed and how it might be manipulated to better control malignancy, infection, and inflammatory diseases.

A major cell extrinsic influence on NKRR formation are ligands encoded by MHC-I genes (1113). The influence of MHC-I on the NKRR presumably reflects education of the developing NK cell to avoid reactivity against self. There are two models that have been proposed for the MHC-specific education of the murine NKRR: the sequential activation model and the two-step selection model (14). The first model proposes that a developing NK cell sequentially acquires or activates Ly49 genes, and it does so in a seemingly random fashion. Once the NK cell expresses a Ly49 gene, expression is maintained throughout the life of that NK cell. The NK cell acquires expression of a sufficient number of self-reactive inhibitory receptors to establish an inhibitory signaling threshold that prevents inappropriate killing of normal host cells. Thus, interaction of inhibitory receptors with host MHC signals the NK cell to terminate further expression of other Ly49 genes and complete the maturation process. In the alternative model, a developing NK cell acquires a fully formed repertoire at an initial stage of development by a stochastic process, but can undergo two possible types of selection. NK cells that have at least one self-specific Ly49 inhibitory receptor are positively selected for while NK cells that express multiple self-specific Ly49 receptors are selected against to avoid accumulation of cells in the compartment with too high of an inhibitory threshold (14).

In addition to influencing formation of the NKRR, MHC-I interactions with self-specific Ly49 or killer immunoglobulin-like receptor (KIR) are also necessary for efficient NK function through a process referred to as licensing (15), disarming (16), or tuning (17). For example, a Ly49A receptor that has high affinity for H-2d molecules can license or tune NK function in H-2d haplotype mice (15). In the absence of these interactions, others argue that the NK cell is disarmed. The recent demonstration that NK cells also express inhibitory and activating SLAM family receptors for ubiquitous self-ligands encoded outside the MHC locus (18), such as CD48, is likely to increase the complexity of NK education by self-ligands and their receptors.

The signal transduction pathways that regulate NKRR formation are incompletely defined. PI3K (19) and the SH2-containing Inositol Phosphatase-1 (SHIP) (3) are recruited to inhibitory NKR upon MHC-I engagement and are therefore able to control activation of Akt/PKB in NK cells. This suggests a role for inositol phospholipid signaling in the regulation of the NKRR by allowing differential turnover of various NK cell subsets (indirect effects) and directly by alteration of NKR gene expression by transcription factors that are distal mediators of PI3K signaling. The former was shown to be the case in SHIP−/− mice, in which a specific NK subset that dominated the compartment also exhibited decreased turnover (3). We have also shown that SHIP functions in NK cells to prevent certain inhibitory receptors from dominating the NKRR (35). We found that on a C57BL/6 background (H-2b) SHIP-deficiency leads to a number of signaling and gene expression perturbations that culminate in an NK cell being hyporesponsive to complex tumor targets that express both MHC-I and the activating ligands, Rae1 or CMV m157 (4, 5). These disruptions include increased expression of 2B4 and the tyrosine phosphatase, SHP1, and inappropriate recruitment and activity of SHP1 at 2B4. This signaling complex creates an inhibitory state at the 2B4 receptor, rendering SHIP−/− NK cells hyporesponsive.

To determine whether 2B4 receptor dominance might also be responsible for the NKRR disruption we observe in SHIP-deficient mice and whether another receptor might dominate cytolytic function in the absence of 2B4, we created 2B4−/−SHIP−/− mice on an H-2b background. As anticipated from our previous studies (4, 5), 2B4 deficiency restores the ability of SHIP−/− NK cells to kill via NKG2D; however, we find that both 2B4 and SHIP are required for formation of a normal NK cell repertoire and IFN-γ induction by key NK activating receptors. Surprisingly, we find that 2B4 inhibits IFN-γ induction by NKp46, but is required for normal induction of IFN-γ by NK1.1. In addition, we demonstrate a novel role for SHIP and 2B4 in preventing NK lineage inappropriate expression of Ly49B. Intriguingly, we also find that the MHC haplotype can overcome the negative impact of 2B4 overexpression on SHIP−/− NK cytolytic function, because deregulated 2B4 expression by SHIP−/− H-2d NK cells does not compromise their ability to lyse MHC-matched targets or MHC-mismatched targets, indicating MHC haplotype has a critical role in determining the cytolytic competency of SHIP-deficient NK cells. Significantly, H-2d NK cells, similar to H-2b NK cells, display a highly disrupted NKR. Thus, the interplay of SHIP, 2B4, and MHC influences NKRR formation, IFN-γ production, and cytolytic function in the NK compartment.

All H-2b NK repertoire analyses described herein are derived from analysis of SHIP+/+ and SHIP−/− mice derived from intercrosses of SHIP+/− mice F10XC57BL6/J mice. SHIP−/−2B4−/− were generated by intercrossing 2B4−/− mice with our SHIP+/− mice (2B4−/− mice were provided by J.D. Schatzle, University of Texas Southwestern Medical Center, Dallas, TX). The 2B4−/−SHIP−/− genotype of the offspring from these matings was confirmed by flow cytometric analysis of viable cells and PCR analysis of genomic DNA. SHIP−/− H-2d mice were generated by crossing SHIP+/− mice to the B10.D2 (H-2d) strain. The progeny of these initial crosses were then backcrossed once more to the H-2d congenic strain to obtain SHIP+/− males and females homozygous for the H-2d haplotypes. H-2d homozygous SHIP−/− males and females were identified, and their SHIP+/− progeny intercrossed to generate wild type (WT) and SHIP−/− progeny for NK repertoire studies on the H-2d haplotypes. All NK repertoire analyses were performed with mice between 6 and 9 wk old.

Anti-CD16/32 was coincubated with the samples to block Fc receptor binding. Abs used for staining included: NK1.1(PK136) (mIgG2a); CD3ε and TCRβ; Ly49A(A1) and Ly49C/I(5E6) (mIgG2a,κ); Ly49F(HBF-719) and Ly49I(YLI-90) (mIgG1,κ); Ly49G2(4D11) and CD94(18d3) (rIgG2a,κ) were obtained from BD Pharmingen (San Jose, CA). 2B4(244F4) (rIgG2a,κ), Ly49H(3D10) (mIgG1), Ly49D(4E5) (rIgG2a,κ), and C7 (hIgG1) were purchased from eBioscience (San Diego, CA). Anti-KLRE1(7E8) (20) (rIgG1), NKRP1D(2D12) (mIgG2a), Ly49B(A1) (rIgG1) (21), KLRG1 Abs (22) were previously described and are conjugated to biotin and revealed with SA-APC as described in this study. Samples were acquired on a FACSCalibur (BD Biosciences, San Jose, CA) and analyzed using FlowJo8 (Tree Star, Ashland OR). Dead cells were excluded from the analysis following cytometer acquisition of staining data based on exclusion of the 7AAD dye.

Cytolysis of RMA, RMA-Rae1+, A20, and BCL1 targets was measured in a standard 4 h 51Cr release assay as previously described (4, 5). Statistics were calculated with Prism software (GraphPad Software, La Jolla, CA) using the Student t test.

To stimulate NK cells, 4–6 million splenocytes from naive mice were incubated with Ab-coated 6-well plates for 5-6 h at 37°C in the presence of GolgiPlug (BD Biosciences). Plates were coated with anti-NK1.1(PK136), anti-NKG2D(A10), or anti-NKp46/NCR1 for 2 h at 37°C. Spleens were harvested and put into single-cell suspension by passing through a 70-μM cell strainer. RBCs were lysed with ACK buffer for 5 min at room temperature. Cells were washed with cold PBS and resuspended in RPMI 1640 supplemented with 10% FBS, 1% penicillin-streptomycin, 1% l-glutamine, 1% sodium pyruvate, and 1% nonessential amino acids. After a 5–6 h incubation, the cells were harvested, Fc receptors were blocked and the cells stained for DX5, TCRβ, and IFN-γ, with the latter stain performed following cell permeabilization and fixation. For H-2d licensing assays, mice were injected i.p. with 70 μg polyinosinic-polycytidylic acid on day –1, and spleens were harvested on day 0.

To assess whether 2B4 expression influences homeostasis, repertoire formation, cytolysis and IFN-γ induction in SHIP-deficient NK cells, we generated 2B4−/−SHIP−/− mice on an H-2b background. 2B4−/−SHIP−/− mice exhibit the same pathologies previously reported in SHIP−/− mice including splenomegaly, weight loss, and a crystalline pneumonia that culminates in their demise at 8–10 wk old. As reported previously (3), we continue to observe a significant increase in peripheral NK cell numbers in SHIP-deficient mice; however, this expansion is dependent on 2B4, because 2B4−/−SHIP−/− mice have normal peripheral NK cell numbers comparable to that of 2B4−/−SHIP+/+ and 2B4+/+SHIP+/+ controls (Fig. 1A).

FIGURE 1.

Defective IFN-γ production in NK cells that lack expression of SHIP, 2B4 or both. A, Representative NK1.1 versus CD3 contour plots of splenocytes from the indicated genotype and bar graph indicating the mean percent of splenic NK cells for each genotype (n = 5 per genotype). B, Representative DX5 versus IFN-γ contour plots (back gated on DX5+TCRβ) of splenocytes exposed to wells coated with anti-NK1.1, anti-NKp46, or anti-NKG2D. Genotypes of the NK cells and the percentage of IFN-γ+ NK cells are indicated. C–E, Percentage of IFN-γ+ NK cells in each genotype in response to NK activating receptor cross-linking NK1.1 (10 μg; C), NKp46 (50 μg; D), and NKG2D (25 μg; E). *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 1.

Defective IFN-γ production in NK cells that lack expression of SHIP, 2B4 or both. A, Representative NK1.1 versus CD3 contour plots of splenocytes from the indicated genotype and bar graph indicating the mean percent of splenic NK cells for each genotype (n = 5 per genotype). B, Representative DX5 versus IFN-γ contour plots (back gated on DX5+TCRβ) of splenocytes exposed to wells coated with anti-NK1.1, anti-NKp46, or anti-NKG2D. Genotypes of the NK cells and the percentage of IFN-γ+ NK cells are indicated. C–E, Percentage of IFN-γ+ NK cells in each genotype in response to NK activating receptor cross-linking NK1.1 (10 μg; C), NKp46 (50 μg; D), and NKG2D (25 μg; E). *p < 0.05; **p < 0.01; ***p < 0.001.

Close modal

The interaction of self-specific NK inhibitory Ly49 receptors with MHC has been shown to endow murine and human NK cells with cytolytic competence, a process alternatively referred to as NK licensing (15) or disarming (16). The functional competence of NK cells can be assessed by intracellular flow cytometric detection of IFN-γ production by freshly isolated splenocytes following Ab mediated cross-linking of an activating receptor such as NK1.1, NKp46, or NKG2D (15). Analysis of the frequency of IFN-γ–producing NK cells following engagement by plate-bound anti-NK1.1, -NKG2D, or -NKp46 revealed that SHIP−/− NK cells have defective IFN-γ production relative to WT controls (Fig. 1A, 1D), which is consistent with their previously reported cytolytic effector function defect (35). However, the absence of 2B4 expression in SHIP-deficient NK cells (2B4−/−SHIP−/− genotype) did not restore IFN-γ induction by any NK-activating receptor tested, including NK1.1, NKp46, and NKG2D (Fig. 1B, 1E). Surprisingly, we find that 2B4 also plays prominent role in induction of IFN-γ by NK-activating receptors, because 2B4−/− NK cells exhibit significantly reduced induction after NK1.1 engagement (Fig. 1B, 1C) and supernormal induction following NKp46 engagement (Fig. 1B, 1D). NKG2D induction of IFN-γ is normal in 2B4−/− NK cells (Fig. 1B, 1E). The supernormal induction of IFN-γ is likely due to the increased frequency of NKp46-expressing cells present in the 2B4−/− NK compartment (see below). However, NK1.1 levels are normal in 2B4−/− NK cells, implying that signals from this CD48 receptor facilitate induction of IFN-γ.

The above results demonstrate that 2B4 deficiency restores normal homeostatic control to the peripheral NK compartment, but does not restore the ability of key NK activating receptors, including NKG2D, to induce IFN-γ production. To determine whether 2B4 could restore cytolytic function, we then assessed the ability of NK cells from 2B4−/−SHIP−/− mice to lyse RMA/Rae1+ targets as compared with NK cells from WT (2B4+/+SHIP+/+), 2B4−/−, and SHIP−/− mice. Cytolysis of RMA/Rae1+ targets was measured in a standard 4 h 51Cr release assay with different E:T ratios, using IL-2–activated NK cells (Fig. 2). In fact, we find that 2B4 deficiency restores SHIP−/− NK killing to WT levels at all E:T ratios tested. Thus, 2B4 expression compromises SHIP−/− NK cytolysis of complex targets that express MHC-I with an NKG2D ligand. However, these results indicate the functional competence of NK cells, as measured by IFN-γ production, is not inextricably linked cytolytic competency.

FIGURE 2.

Inhibitory dominance by 2B4 in BL6 SHIP−/− NK cells. NK cells were magnetically enriched from splenocytes of mice of the indicated genotypes by AutoMACS depletion of B, T, and myeloid cells and then cultured in 2000 U/ml of human recombinant IL-2 for 5–7 d. Cytolysis of RMA/Rae1+ transfectants was then assessed by a 4 h 51Cr release assay at the indicated E:T ratios. Percent % Lysis = 100 × [(Experimental release − Spontaneous release)/(Maximum release − Spontaneous release)]. This analysis shows that 2B4 deficiency restores SHIP−/− NK killing to WT levels at all E:T ratios tested (p < 0.05, using Prism software and calculating using the Student t test). These studies are representative of three independent experiments. Killing for all three genotypes was significantly greater than for SHIP−/− at all E:T ratios tested (*p < 0.05).

FIGURE 2.

Inhibitory dominance by 2B4 in BL6 SHIP−/− NK cells. NK cells were magnetically enriched from splenocytes of mice of the indicated genotypes by AutoMACS depletion of B, T, and myeloid cells and then cultured in 2000 U/ml of human recombinant IL-2 for 5–7 d. Cytolysis of RMA/Rae1+ transfectants was then assessed by a 4 h 51Cr release assay at the indicated E:T ratios. Percent % Lysis = 100 × [(Experimental release − Spontaneous release)/(Maximum release − Spontaneous release)]. This analysis shows that 2B4 deficiency restores SHIP−/− NK killing to WT levels at all E:T ratios tested (p < 0.05, using Prism software and calculating using the Student t test). These studies are representative of three independent experiments. Killing for all three genotypes was significantly greater than for SHIP−/− at all E:T ratios tested (*p < 0.05).

Close modal

SHIP-deficient NK cells have a highly disrupted NKRR, with underrepresentation of many NK receptors and overexpression of 2B4 (Fig. 3B) (35). We then asked whether 2B4 deficiency might also restore formation of a normal receptor repertoire in the peripheral NK compartment of SHIP-deficient mice. To determine how SHIP and 2B4 contribute to the acquisition of the NKRR, we performed flow cytometric analysis on splenocytes from naive mice, first gating on NK1.1+Lin cells and then analyzing the percent expression of the indicated receptors (Fig. 3). These data are depicted as bar graphs where the representation of each NK receptor in the compartment is normalized to that of WT controls analyzed in parallel with each of the three mutant genotypes (2B4−/−SHIP+/+, SHIP−/−2B4−/−, SHIP−/−2B4+/+; Fig. 3A, 3C). With the exception of 2B4, all “% of normal” values refer to the frequency of NK cells in the indicated mutant after normalization to WT. The 2B4 “% of normal” values refer to mean fluorescence intensity (MFI) or surface density after normalization to WT controls.

FIGURE 3.

Normalized representation of each NKR in the peripheral NK compartment of 2B4−/−SHIP−/− (Α), SHIP−/− (B), and 2B4−/− mice (C). All percent and MFI for NK receptors were determined after gating on NK1.1+Lin splenocytes of 6–8-wk-old adult mice (Lin panel: IgM, CD3, TcR-β, Gr1, CD11c). To estimate the percentage of NKR+ cells in the NK compartment, positive NKR gates were set at ≥95% of NK1.1+Lin cells staining positive for an isotype control stain performed on an equal mixture of null and WT splenocytes. Representation of individual NKRs in splenic SHIP−/−, SHIP−/−2B4−/−, and 2B4−/− NK cells is presented after normalization to WT. The percent of normal was calculated as follows: (%NKR+ genotype X/%NKR+ SHIP+/+) × 100, for each indicated NKR, where X represents the genotypes indicated (e.g., SHIP, 2B4, or SHIP 2B4 double knockout). For the 2B4 receptor, percent normal was calculated in the same manner, except that MFI was used rather than %NKR+. White, black, and gray bar graphs represent percent normal values that are significantly lower, higher, or unchanged in the SHIP−/−, SHIP−/−2B4−/−, and 2B4−/− NK compartment as compared with WT, respectively. *p < 0.05.

FIGURE 3.

Normalized representation of each NKR in the peripheral NK compartment of 2B4−/−SHIP−/− (Α), SHIP−/− (B), and 2B4−/− mice (C). All percent and MFI for NK receptors were determined after gating on NK1.1+Lin splenocytes of 6–8-wk-old adult mice (Lin panel: IgM, CD3, TcR-β, Gr1, CD11c). To estimate the percentage of NKR+ cells in the NK compartment, positive NKR gates were set at ≥95% of NK1.1+Lin cells staining positive for an isotype control stain performed on an equal mixture of null and WT splenocytes. Representation of individual NKRs in splenic SHIP−/−, SHIP−/−2B4−/−, and 2B4−/− NK cells is presented after normalization to WT. The percent of normal was calculated as follows: (%NKR+ genotype X/%NKR+ SHIP+/+) × 100, for each indicated NKR, where X represents the genotypes indicated (e.g., SHIP, 2B4, or SHIP 2B4 double knockout). For the 2B4 receptor, percent normal was calculated in the same manner, except that MFI was used rather than %NKR+. White, black, and gray bar graphs represent percent normal values that are significantly lower, higher, or unchanged in the SHIP−/−, SHIP−/−2B4−/−, and 2B4−/− NK compartment as compared with WT, respectively. *p < 0.05.

Close modal

Contrary to what was observed with cytolytic function, we find that 2B4 deficiency does not restore repertoire formation, because the repertoire of 2B4−/−SHIP−/− NK cells is severely disrupted with 12 of 14 NK receptors having a significantly abnormal representation (Fig. 3A). We also observe a significant degree of repertoire disruption in 2B4−/− mice with 9 of 14 NK receptors significantly altered as compared with WT controls (Fig. 3C). Thus, expression of both 2B4 and SHIP is required for the normal development of the NKRR. Because SHIP is recruited to 2B4 in NK cells (23) and can influence the role of 2B4 role in cytolytic function (4, 5), it is possible that 2B4 and SHIP interact in a signaling pathway that also promotes NK repertoire formation. Thus, NK receptors whose expression is altered in a similar manner by 2B4, SHIP, or combined 2B4/SHIP deficiency are then potentially regulated by signaling pathways controlled by a 2B4:SHIP complex. However, certain NK receptors exhibit different patterns of disruption in 2B4- and SHIP-deficient NK cells (Fig. 3B, 3C). For example, 2B4 deficiency leads to increased representation of key activating receptors such as NKp46 and DNAM-1 in the NK compartment, but only in the context of SHIP competency (Fig. 3A). On the contrary, there are significantly fewer SHIP-deficient NK cells that express NKp46 and NKG2D (Fig 3B). Thus, 2B4 and SHIP have overlapping and distinct effects on the NKRR. We find that 2B4 deficiency restores cytolytic function and homeostatic control to SHIP-deficient NK cells, but is unable to restore normal repertoire formation or IFN-γ induction.

In addition to the NK-associated receptors analyzed above, we also examined the role of SHIP and 2B4 on Ly49B expression. Ly49B and Ly49Q are the only members of the Ly49 gene family not expressed by NK cells, but instead are restricted to myeloid lineage cells (21). Like the Ly49A and C receptors we previously found to be regulated by SHIP, Ly49B is also a promiscuous MHC-I receptor that does not exhibit precise specificity for a given MHC haplotype (24). SHIP is also recruited to Ly49B (21) making it a strong candidate for a SHIP-regulated receptor. Analysis of Ly49B expression revealed that it is expressed on a significant proportion of SHIP−/− NK cells in either an H-2b (C57BL/6) background (Fig. 4A) or an H-2d (B10.D2) background (Fig. 4B) in both the spleen and bone marrow (BM), whereas virtually no Ly49B expression was detected on WT NK cells from both haplotypes, consistent with the findings of Gays et al. (21). We performed Ly49B blocking studies using hybridoma supernatants or purified Ab against Ly49B (21) in H-2d SHIP−/− and WT 51Cr release assays. Blocking of Ly49B did not significantly increase or decrease cytolysis by SHIP−/− LAK cells (data not shown). We also find that IFN-γ induction by NK1.1 and NKG2D cross-linking is not significantly altered when Ly49B is blocked (data not shown). These in vitro assays indicate that deregulated Ly49B expression may not influence SHIP−/− NK effector functions. Whether there are functional consequences for NK cells in vivo that arise owing to lineage-inappropriate expression of Ly49B will require the development of Ly49B−/− mice and thus remain to be determined. 2B4−/− NK cells from the BM and spleen also express Ly49B at significant levels (Fig. 4C, 4E, 4F). This finding suggests a direct role for 2B4 in lineage-specific regulation of Ly49 receptors. Not surprisingly, BM and splenic NK cells from SHIP−/−2B4−/− mice also express high levels of Ly49B (Fig. 4D4F).

FIGURE 4.

SHIP and 2B4 prevent lineage-inappropriate expression of Ly49B. Spleen and bone marrow of H-2b mice (A) and H-2d mice (B). Gates for Ly49B+ in SHIP−/− (black histogram) or SHIP+/+ (gray histogram) were based on gating beginning at the 95th percentile of a rIgG1 isotype control. Statistical analysis of the frequency of Ly49B expression in H-2b and H-2d SHIP−/− versus SHIP+/+ mice. *p < 0.05, as determined by Student t test. Ly49B staining on BM and splenic NK1.1+CD3 NK cells from 2B4−/− (C) and 2B4−/−SHIP−/− mice (D). Black histograms are Ly49B staining, and gray histograms are isotype control stains. Statistical analysis of the frequency of Ly49B expression in the BM (E) and spleens (F) in the following three genotypes: 2B4+/+SHIP+/+ (WT), 2B4−/−SHIP−/−, and 2B4−/−SHIP+/+. *p < 0.05; ***p < 0.001.

FIGURE 4.

SHIP and 2B4 prevent lineage-inappropriate expression of Ly49B. Spleen and bone marrow of H-2b mice (A) and H-2d mice (B). Gates for Ly49B+ in SHIP−/− (black histogram) or SHIP+/+ (gray histogram) were based on gating beginning at the 95th percentile of a rIgG1 isotype control. Statistical analysis of the frequency of Ly49B expression in H-2b and H-2d SHIP−/− versus SHIP+/+ mice. *p < 0.05, as determined by Student t test. Ly49B staining on BM and splenic NK1.1+CD3 NK cells from 2B4−/− (C) and 2B4−/−SHIP−/− mice (D). Black histograms are Ly49B staining, and gray histograms are isotype control stains. Statistical analysis of the frequency of Ly49B expression in the BM (E) and spleens (F) in the following three genotypes: 2B4+/+SHIP+/+ (WT), 2B4−/−SHIP−/−, and 2B4−/−SHIP+/+. *p < 0.05; ***p < 0.001.

Close modal

As with 2B4, we find that the expression or representation of most Ly49 receptors is abnormal in the peripheral NK compartment of H-2b SHIP−/− mice (Fig. 3). To determine whether variation in MHC haplotype might alter how the repertoire is disrupted in SHIP−/− NK cells, we assessed the NK repertoire perturbation in SHIP−/− mice with an H-2d haplotype known to have a unique effect on the representation of certain Ly49 receptors. For example, surface expression of Ly49A is downmodulated in the presence of high-affinity MHC-I ligands in the H-2d locus (12). Analysis of the peripheral NK repertoire in SHIP−/− mice with an H-2d haplotype (Fig. 5A) indicated that the repertoire perturbation in these mice is essentially identical to that of SHIP−/− H-2b mice, with the exception of Ly49A, NKp46, and DNAM-1. We find that Ly49A is significantly overexpressed in H-2d SHIP−/− NK cells relative to NK cells of WT littermates (Fig. 5A, 5B). Consistent with the MHC independence of its ligand, the surface density of 2B4 is abnormally high in the peripheral NK compartment of SHIP−/− H-2d mice, similar to what we observed on an H-2b background. The frequency of NKp46+ NK cells is not significantly reduced as it is in H2b SHIP−/− NK cells (Fig. 5A), whereas DNAM-1 is highly overexpressed in H2d+ SHIP−/− NK cells as compared with H2d+ WT controls (Fig. 5A).

FIGURE 5.

2B4 is not a dominant inhibitory receptor in H-2d SHIP−/− NK cells where a strong licensing receptor is overexpressed. A, Normalized representation of the NKRR in SHIP−/− splenic H-2d NK cells (n = 6 mice for each genotype). Representation of individual NKRs in splenic SHIP−/− NK cells are presented after normalization to WT. The percent of normal = (%NKR+ SHIP−/−/%NKR+ SHIP+/+) × 100, for each indicated NKR. For Ly49A and 2B4 receptors, percent normal was calculated in the same manner, except that MFI was used rather than %NKR+. White, black, and gray bar graphs represent percent normal values that are significantly lower, higher, or unchanged in the SHIP−/− NK compartment as compared with WT, respectively. *p < 0.05. B, Representative Ly49A staining on NK1.1+CD3 splenic NK cells on WT and SHIP−/− H-2d backgrounds. C, SHIP−/− and WT NK cytolysis of BCL1 (H-2d) lymphoma targets, A20 (H-2d) lymphoma targets (D), and RMA (H-2b) lymphoma targets (E). Cytolysis was analyzed in a standard 4 h 51Cr release assay. Each analysis is representative of at least two independent experiments. *p < 0.05.

FIGURE 5.

2B4 is not a dominant inhibitory receptor in H-2d SHIP−/− NK cells where a strong licensing receptor is overexpressed. A, Normalized representation of the NKRR in SHIP−/− splenic H-2d NK cells (n = 6 mice for each genotype). Representation of individual NKRs in splenic SHIP−/− NK cells are presented after normalization to WT. The percent of normal = (%NKR+ SHIP−/−/%NKR+ SHIP+/+) × 100, for each indicated NKR. For Ly49A and 2B4 receptors, percent normal was calculated in the same manner, except that MFI was used rather than %NKR+. White, black, and gray bar graphs represent percent normal values that are significantly lower, higher, or unchanged in the SHIP−/− NK compartment as compared with WT, respectively. *p < 0.05. B, Representative Ly49A staining on NK1.1+CD3 splenic NK cells on WT and SHIP−/− H-2d backgrounds. C, SHIP−/− and WT NK cytolysis of BCL1 (H-2d) lymphoma targets, A20 (H-2d) lymphoma targets (D), and RMA (H-2b) lymphoma targets (E). Cytolysis was analyzed in a standard 4 h 51Cr release assay. Each analysis is representative of at least two independent experiments. *p < 0.05.

Close modal

As mentioned above, Ly49A has been shown to be downmodulated in the presence of H-2d (12). According to a newly proposed quantitative regulation model of NK education (17), high expression of Ly49A by NK cells in an MHC-I background with a strong educating impact should reduce the threshold required for NK cell activation. To assess whether inappropriate expression of 2B4 or Ly49A might alter cytolytic function in H-2d SHIP−/− NK cells, we compared the ability of SHIP−/− and WT H-2d NK cells to kill H-2d and H-2b tumor targets. We find that SHIP−/− H-2d NK cells have normal cytolytic activity against two different H-2d+ MHC-I matched tumor targets, BCL1 and A20 (Fig. 5C, 5D). We find that overexpression of 2B4 (Fig. 5A) does not compromise H-2d SHIP−/− NK cytolytic function as it does on an H-2b background. Thus, Ly49A overexpression by SHIP−/− H-2d NK cells does not function as a dominant inhibitory receptor and impair cytolysis. In fact, we find that H-2d SHIP−/− NK cells exhibit supernormal cytolytic activity against MHC mismatched H-2b tumor targets, RMA cells (Fig. 5E). However, this finding is not the case for SHIP−/− NK cells with an H-2b haplotype, because these cells exhibit impaired cytolytic function against the MHC-I mismatched targets A20 and BCL-1 (data not shown). The increased expression of the DNAM-1 activating receptor in SHIP−/− H-2d NK cells could potentially account for their normal or enhanced cytolytic function against MHC-matched and mismatched targets, respectively. Thus, MHC haplotype can modulate the effect of SHIP on the NK repertoire and its impact on effector function, such that SHIP-deficiency in certain MHC-I mismatched contexts can promote supernormal cytolytic activity by NK cells.

Our analysis of 2B4−/−SHIP−/− NK cells indicated discordance in cytolytic function and IFN-γ induction. We then wanted to determine whether the normal or supernormal cytolytic capacity of H2d+ SHIP−/− NK cells could also result in the restoration of normal or above-normal IFN-γ induction from key NK activating receptors. To investigate this, we primed the mice with polyinosinic-polycytidylic acid (day −1) and harvested the spleens on day 0. To induce IFN-γ production, we cross-linked NK1.1 (Fig. 6A, 6C) or NKG2D (Fig. 6A, 6D) using plate-bound mAbs. We measured the production of IFN-γ by intracellular flow, and we show here that SHIP-deficient NK cells from H-2d mice have a markedly impaired ability to induce IFN-γ from both NK activating receptors (Fig. 6A–D). This finding is consistent with the discordance of cytolytic function and IFN-γ induction in 2B4−/−SHIP−/− NK cells, and it indicates that although in certain genetic contexts (2B4 deficiency, H2d hapolotype) SHIP-deficient NK cells can have normal or supernormal cytolytic function, they nonetheless remain poor producers of IFN-γ in response to engagement of major NK activating receptors. We can conclude that SHIP expression is a uniform and essential requirement for this NK effector function.

FIGURE 6.

H-2d NK cells exhibit an impaired ability to produce IFN-γ. Representative DX5 versus IFN-γ contour plots (back gated on DX5+TCRβ-) of spleen cells (A) unstimulated (PBS) or stimulated with 50 μg anti-NK1.1 or 50 μg anti-NKG2D. B–D, Statistical analysis of licensed NK cells from SHIP+/+ and SHIP−/− mice with an H-2d haplotype. ***p < 0.0001, compared with WT control. Representative of at least three independent experiments.

FIGURE 6.

H-2d NK cells exhibit an impaired ability to produce IFN-γ. Representative DX5 versus IFN-γ contour plots (back gated on DX5+TCRβ-) of spleen cells (A) unstimulated (PBS) or stimulated with 50 μg anti-NK1.1 or 50 μg anti-NKG2D. B–D, Statistical analysis of licensed NK cells from SHIP+/+ and SHIP−/− mice with an H-2d haplotype. ***p < 0.0001, compared with WT control. Representative of at least three independent experiments.

Close modal

In this study, we provide genetic evidence that the interaction of SHIP with both 2B4 and MHC-I loci is required for the acquisition of a normal repertoire of inhibitory and activating receptors, normal cytolytic function, and induction of IFN-γ production by key NK activating receptors. Consistent with our previous study showing that blockade of CD48 on complex MHC-I+Rae1+ targets restored normal cytolysis to SHIP-deficient NK cells (4), the cytolytic defect of SHIP-deficient NK cells is corrected by 2B4 deficiency. Cytolytic competence against Rae1+ target cells by 2B4−/−SHIP−/− and 2B4−/− NK cells occurs despite a highly disrupted NKRR. However, SHIP−/−2B4−/− and H-2d SHIP−/− NK cells exhibit defective induction of IFN-γ, although they exhibit normal or supernormal cytolytic capacity. This defect in IFN-γ induction can result from decreased expression of NKp46 and NKG2D in SHIP−/− NK cells; however, NK1.1 receptor expression levels are normal or elevated on SHIP−/− NK cells (3), indicating that SHIP actually is required downstream of certain NK activating receptors for efficient induction of IFN-γ. Surprisingly, we find that 2B4 restrains induction of IFN-γ in response to engagement of the NK activating receptor NKp46, but not NK1.1. These findings demonstrate that IFN-γ induction and cytolytic competence are regulated by distinct mechanisms in NK cells and that both SHIP and 2B4 play a prominent role in IFN-γ production by NK cells by modulating activating receptor expression and/or signaling pathways downstream of these receptors.

The finding that 2B4 limits the expression of key activating receptors, NKp46 and DNAM-1, is attractive and may explain why we and Vaidya et al. (25) observed enhanced killing by 2B4−/− NK cells. These data suggest that 2B4 may be necessary to limit the expression of these activating receptors. In addition, we find that SHIP and 2B4 are required to prevent the lineage-inappropriate expression of Ly49B, the polyspecific MHC-I receptor that is normally restricted to myeloid cells (21). This finding establishes a previously unappreciated role for SHIP and 2B4 in maintaining lineage restricted expression of immune receptors. The in vivo functional consequences of inappropriate Ly49B expression in the NK cell compartment of SHIP−/−, SHIP−/−2B4−/−, and 2B4−/− mice remains unknown, but certainly merits further investigation.

We also find that the effects of SHIP on repertoire formation and cytolytic function is influenced by the composition of MHC-I ligands, because a potent educating or licensing receptor, Ly49A, is overexpressed by SHIP-deficient NK cells in the presence of its high-affinity H-2d ligand. Surprisingly, cytolytic function is not found to be defective in SHIP−/− H-2d NK cells, suggesting that the increased educating or licensing capacity of Ly49A in H-2d+ SHIP−/− NK cells could counteract inhibitory signals resulting from overexpression of 2B4; however, H-2d NK cells have defective IFN-γ induction. The normal or supernormal cytolytic activity we observe with H-2d SHIP−/− NK cells could also be due to the normal levels of NKp46 expression or increased expression of DNAM-1, respectively, that we observe in H-2d SHIP−/− NK cells. Based on our findings, we suggest that the IFN-γ induction assay may not always be a valid surrogate for NK cytolytic competence, particularly for analysis of signal transduction mutations that can effect NK function. Others have also observed a similar discordance of IFN-γ induction and cytolysis in Bcl10−/− NK cells (26).

The interaction of SHIP and 2B4 influences signaling pathways that determine the cytolytic function of mature NK cells in both humans and mice (5, 23). Our findings suggest that 2B4:SHIP signaling could also play a role in early NK development to promote the efficient acquisition of NK receptors that sense MHC-I ligands. In fact, expression of 2B4 (27, 28) and SHIP precedes expression of the Ly49 and CD94/NKG2 receptors in NK development. The early expression of 2B4 in developing NK cells may be necessary to achieve self-tolerance until a properly diverse Ly49 and CD94/NKG2 repertoire is acquired. Consistent with this hypothesis, McNerney et al. (29) found that 2B4 promotes self tolerance by mature murine NK cells in rodents, and Sivori et al. (30) showed that 2B4 inhibitory signals limits cytolysis by NK cells that have yet to acquire KIR expression. This putative function of 2B4 in developing NK cells could be particularly important in preventing inappropriate NK cytolysis in the BM compartment, because immature NK cells have been shown to acquire cytolytic activity prior to acquisition of Ly49 receptors and KIR receptors (27, 30). Consistent with this hypothesis, the ligand for 2B4, CD48, is ubiquitously expressed in the developing hematopoietic system providing nearly constant interaction of developing NK cells with a tolerizing signal in the form of CD48. The exception to this is the primitive subset of hematopoietic stem cells (HSCs) that lack CD48 expression (31). Presumably developing NK cells do not co-occupy the endosteal niche where primitive HSC reside, although this merits direct analysis.

A role for 2B4 in maintaining self-tolerance in early NK cells and a lack of CD48 expression by HSCs would enable the NK lineage to modulate hematopoiesis by lysis of CD48 HSCs. Because NK cells are the only cytolytic lymphocyte that develops in the BM, this role for 2B4 would provide a means for NK cells to mediate negative feedback on HSCs and thus lymphoid output. The plausibility of such a mechanism is further suggested by evidence that syngeneic HSC function can be limited by NK cells in vivo (32). Thus, a 2B4:SHIP complex could potentially play a role in such a lymphoid feedback pathway. Consistent with this hypothesis, CD48 HSCs inappropriately accumulate in the BM of SHIP−/− mice (33). An additional role for the 2B4:SHIP complex in NK tolerance toward APCs in secondary lymphoid tissues is also a distinct possibility. Evidence for this role includes the inappropriate expansion of the dendritic cell compartment in the lymph nodes (LNs) of SHIP−/− mice (34), increased expression of SHIP in a subset of LN NK cells that lack KIR, but express 2B4 (35), and that 2B4 has inhibitory function in human LN NK cells (30).

How 2B4 signals promote NK tolerance toward CD48+ targets in the absence of MHC-I inhibitory receptors, and trigger the acquisition of a full repertoire of MHC-I receptors, remains to be defined. 2B4 can directly or indirectly recruit a wide variety of signaling molecules via its four immunoreceptor tyrosine-based switch motifs. To date, the following signaling components have been shown to be recruited to 2B4: PLC-γ, LAT, Grb2, SAP, Fyn, EAT2, PI3K, SHP1, SHP2, and SHIP (5, 23, 36, 37). Thus, signaling complexes at 2B4 have the biochemical capacity to impact a wide variety of distal signaling pathways that can control NK cell survival, proliferation, and/or gene expression. Because 2B4 is recruited to the NK synapse and SHIP can attenuate PI3K activity from other receptors in trans (38), a 2B4:SHIP complex also has the potential to limit PI3K activity originating at other NK receptors. In fact, immature KIR NK cells can kill autologous cells, including APCs, via PI3K-mediated pathways downstream of NKp46 and NKp30 (30, 39), and 2B4 limits this activity (30). Thus, trans activity of SHIP at 2B4 to oppose PI3K activity at other receptors could be an essential feature of NK tolerance signaling. The expansion of HSCs in BM and APCs in the LNs of SHIP−/− mice is consistent with this possibility.

The NKRR is regulated by elements intrinsic and extrinsic to the NK lineage. Several studies have shown that external signals received by NK cells from self-ligands, such as MHC-I, can influence the repertoire through differential effects on NK subset survival and/or proliferation (40, 41). In addition, NK intrinsic signaling pathways influence NKR expression, including 2B4, via cis-acting sequences present in receptor promoters (7, 8, 42, 43) and by activation of transcription factors that bind to these sites (44, 45). The challenge for biologists studying NK cells is to better understand how integration of these extrinsic and intrinsic pathways determines the final composition of the NKRR. Coexpression of 2B4 and SHIP prior to MHC-I receptor expression suggests that 2B4:SHIP complexes are uniquely positioned to play such a role. For example, the trans activity of SHIP from 2B4 could oppose PI3K activity at MHC-I receptors and thus limit the survival or proliferation of NK subsets expressing these MHC-I receptors. Consistent with this possibility, PI3K can be recruited to MHC-I inhibitory receptors to activate Akt (3, 19). SHIP acting in trans from 2B4 or in cis from the same MHC-I receptors could limit PI3K/Akt survival signals, and thereby prevent inappropriate expansion of such NK subsets. SHIP can also be recruited directly to MHC-I receptors; therefore, SHIP may also limit these subsets in cis (3, 21, 24). This cis activity may be important when MHC-I receptors are in a genetic background where high-affinity MHC-I ligands are also present, as suggested by the overexpression of Ly49A in SHIP−/− H-2d mice.

In addition to limiting expansion of specific subsets of NK cells, it is also possible that 2B4:SHIP complexes influence intrinsic pathways that determine NK receptor expression and effector function. For example, activation of transcription factors known to act on promoters for NK receptors and/or 2B4 (e.g., Ets, NF-κB, CREB) is influenced by 2B4 engagement (43) and signaling molecules recruited to 2B4 (e.g., Fyn, SHIP, PI3K) (46). Consistent with this hypothesis, 2B4 expression is deregulated in SHIP−/− NK cells (4), whereas Fyn−/−, SHIP−/−, and PI3K−/− mutants all exhibit profound disruptions of their MHC-I NK receptor repertoires (3, 9, 47). The receptor expression changes created by SHIP or 2B4 deficiency are likely to cause some of the alterations in effector function that we observe in SHIP−/− and 2B4−/− NK cells. However, independent of these receptor expression changes, 2B4:SHIP complexes also appear to influence signaling pathways that promote NK effector functions. For example, SHIP−/− NK cells are unable to trigger IFN-γ induction in response to NK1.1 engagement despite normal or increased surface density of NK1.1. This finding suggests a role for SHIP in promoting IFN-γ expression via generation of its product PI(3,4)P2, which along with PI(3,4,5)P3 is a critical second messenger for the PI3K pathway. This is consistent with a recent report showing that SHIP promotes, rather than inhibits, macrophage effector function via generation of PI(3,4)P2 (48). Thus, the uniform defect in IFN-γ induction that we observed for SHIP-deficient NK cells, whether 2B4-deficient or of different MHC haplotypes, demonstrates an absolute requirement for SHIP in the induction of IFN-γ production by major NK activating receptors. These findings reveal a pivotal role for the interaction of SHIP and 2B4 in the regulation of the NKRR, cytolytic function, and IFN-γ production.

We thank Nathan Watts, Najwa Khan, Elizabeth Gengo, and Ashley Bekolay for genotyping of mice; Sarah Highfill for technical assistance in the early phase of this study; and Wayne Yokoyama, James Carlyle, and Laurent Brossay the gifts of Ly4H, NKG2D, NKR-P1D, and KLRG1 Abs.

Disclosures The authors have no financial conflicts on interest.

This work was supported in part by grants from the National Institutes of Health (RO1 HL72523 and R01 101748 to W.G.K.) and the Paige Arnold Butterfly Run. W.G.K. was also the Newman Family Scholar of the Leukemia and Lymphoma Society during a portion of this study and is currently the Murphy Family Professor of Children’s Oncology Research and an Empire Scholar at State University of New York Upstate.

Abbreviations used in this paper:

BM

bone marrow

HSC

hematopoietic stem cell

KIR

killer immunoglobulin-like receptor

LN

lymph node

MFI

mean fluorescence intensity

MHC-I

MHC class I

NKR

natural killer receptor

NKRR

natural killer receptor repertoire

WT

wild type.

1
Lanier
L. L.
1998
.
NK cell receptors.
Annu. Rev. Immunol.
16
:
359
393
.
2
Gasser
S.
,
Orsulic
S.
,
Brown
E. J.
,
Raulet
D. H.
.
2005
.
The DNA damage pathway regulates innate immune system ligands of the NKG2D receptor.
Nature
436
:
1186
1190
.
3
Wang
J. W.
,
Howson
J. M.
,
Ghansah
T.
,
Desponts
C.
,
Ninos
J. M.
,
May
S. L.
,
Nguyen
K. H.
,
Toyama-Sorimachi
N.
,
Kerr
W. G.
.
2002
.
Influence of SHIP on the NK repertoire and allogeneic bone marrow transplantation.
Science
295
:
2094
2097
.
4
Wahle
J. A.
,
Paraiso
K. H.
,
Costello
A. L.
,
Goll
E. L.
,
Sentman
C. L.
,
Kerr
W. G.
.
2006
.
Cutting edge: dominance by an MHC-independent inhibitory receptor compromises NK killing of complex targets.
J. Immunol.
176
:
7165
7169
.
5
Wahle
J. A.
,
Paraiso
K. H.
,
Kendig
R. D.
,
Lawrence
H. R.
,
Chen
L.
,
Wu
J.
,
Kerr
W. G.
.
2007
.
Inappropriate recruitment and activity by the Src homology region 2 domain-containing phosphatase 1 (SHP1) is responsible for receptor dominance in the SHIP-deficient NK cell.
J. Immunol.
179
:
8009
8015
.
6
Kerr
W. G.
2007
.
Acquisition and remodeling of the NK receptor repertoire
. In
Everything You Wanted to Know About NK Cells, But Were Afraid to Ask.
Brossay
L.
, ed,
Research Signpost, Kerala, India. p. 45–58.
7
Saleh
A.
,
Davies
G. E.
,
Pascal
V.
,
Wright
P. W.
,
Hodge
D. L.
,
Cho
E. H.
,
Lockett
S. J.
,
Abshari
M.
,
Anderson
S. K.
.
2004
.
Identification of probabilistic transcriptional switches in the Ly49 gene cluster: a eukaryotic mechanism for selective gene activation.
Immunity
21
:
55
66
.
8
Tanamachi
D. M.
,
Moniot
D. C.
,
Cado
D.
,
Liu
S. D.
,
Hsia
J. K.
,
Raulet
D. H.
.
2004
.
Genomic Ly49A transgenes: basis of variegated Ly49A gene expression and identification of a critical regulatory element.
J. Immunol.
172
:
1074
1082
.
9
Lowin-Kropf
B.
,
Kunz
B.
,
Schneider
P.
,
Held
W.
.
2002
.
A role for the src family kinase Fyn in NK cell activation and the formation of the repertoire of Ly49 receptors.
Eur. J. Immunol.
32
:
773
782
.
10
Tassi
I.
,
Presti
R.
,
Kim
S.
,
Yokoyama
W. M.
,
Gilfillan
S.
,
Colonna
M.
.
2005
.
Phospholipase C-gamma 2 is a critical signaling mediator for murine NK cell activating receptors.
J. Immunol.
175
:
749
754
.
11
Held
W.
,
Cado
D.
,
Raulet
D. H.
.
1996
.
Transgenic expression of the Ly49A natural killer cell receptor confers class I major histocompatibility complex (MHC)-specific inhibition and prevents bone marrow allograft rejection.
J. Exp. Med.
184
:
2037
2041
.
12
Kåse
A.
,
Johansson
M. H.
,
Olsson-Alheim
M. Y.
,
Kärre
K.
,
Höglund
P.
.
1998
.
External and internal calibration of the MHC class I-specific receptor Ly49A on murine natural killer cells.
J. Immunol.
161
:
6133
6138
.
13
Yawata
M.
,
Yawata
N.
,
Draghi
M.
,
Little
A. M.
,
Partheniou
F.
,
Parham
P.
.
2006
.
Roles for HLA and KIR polymorphisms in natural killer cell repertoire selection and modulation of effector function.
J. Exp. Med.
203
:
633
645
.
14
Raulet
D. H.
,
Held
W.
,
Correa
I.
,
Dorfman
J. R.
,
Wu
M. F.
,
Corral
L.
.
1997
.
Specificity, tolerance and developmental regulation of natural killer cells defined by expression of class I-specific Ly49 receptors.
Immunol. Rev.
155
:
41
52
.
15
Kim
S.
,
Poursine-Laurent
J.
,
Truscott
S. M.
,
Lybarger
L.
,
Song
Y. J.
,
Yang
L.
,
French
A. R.
,
Sunwoo
J. B.
,
Lemieux
S.
,
Hansen
T. H.
,
Yokoyama
W. M.
.
2005
.
Licensing of natural killer cells by host major histocompatibility complex class I molecules.
Nature
436
:
709
713
.
16
Raulet
D. H.
,
Vance
R. E.
.
2006
.
Self-tolerance of natural killer cells.
Nat. Rev. Immunol.
6
:
520
531
.
17
Brodin
P.
,
Lakshmikanth
T.
,
Johansson
S.
,
Karre
K.
,
Hoglund
P.
.
2008
.
The strength of inhibitory input during education quantitatively tunes the functional responsiveness of individual natural killer cells.
Blood
113
:
2434
2441
.
18
Lee
K. M.
,
McNerney
M. E.
,
Stepp
S. E.
,
Mathew
P. A.
,
Schatzle
J. D.
,
Bennett
M.
,
Kumar
V.
.
2004
.
2B4 acts as a non-major histocompatibility complex binding inhibitory receptor on mouse natural killer cells.
J. Exp. Med.
199
:
1245
1254
.
19
Marti
F.
,
Xu
C. W.
,
Selvakumar
A.
,
Brent
R.
,
Dupont
B.
,
King
P. D.
.
1998
.
LCK-phosphorylated human killer cell-inhibitory receptors recruit and activate phosphatidylinositol 3-kinase.
Proc. Natl. Acad. Sci. USA
95
:
11810
11815
.
20
Koike
J.
,
Wakao
H.
,
Ishizuka
Y.
,
Sato
T. A.
,
Hamaoki
M.
,
Seino
K.
,
Koseki
H.
,
Nakayama
T.
,
Taniguchi
M.
.
2004
.
Bone marrow allograft rejection mediated by a novel murine NK receptor, NKG2I.
J. Exp. Med.
199
:
137
144
.
21
Gays
F.
,
Aust
J. G.
,
Reid
D. M.
,
Falconer
J.
,
Toyama-Sorimachi
N.
,
Taylor
P. R.
,
Brooks
C. G.
.
2006
.
Ly49B is expressed on multiple subpopulations of myeloid cells.
J. Immunol.
177
:
5840
5851
.
22
Tessmer
M. S.
,
Fugere
C.
,
Stevenaert
F.
,
Naidenko
O. V.
,
Chong
H. J.
,
Leclercq
G.
,
Brossay
L.
.
2007
.
KLRG1 binds cadherins and preferentially associates with SHIP-1.
Int. Immunol.
19
:
391
400
.
23
Eissmann
P.
,
Beauchamp
L.
,
Wooters
J.
,
Tilton
J. C.
,
Long
E. O.
,
Watzl
C.
.
2005
.
Molecular basis for positive and negative signaling by the natural killer cell receptor 2B4 (CD244).
Blood
105
:
4722
4729
.
24
Scarpellino
L.
,
Oeschger
F.
,
Guillaume
P.
,
Coudert
J. D.
,
Lévy
F.
,
Leclercq
G.
,
Held
W.
.
2007
.
Interactions of Ly49 family receptors with MHC class I ligands in trans and cis.
J. Immunol.
178
:
1277
1284
.
25
Vaidya
S. V.
,
Stepp
S. E.
,
McNerney
M. E.
,
Lee
J. K.
,
Bennett
M.
,
Lee
K. M.
,
Stewart
C. L.
,
Kumar
V.
,
Mathew
P. A.
.
2005
.
Targeted disruption of the 2B4 gene in mice reveals an in vivo role of 2B4 (CD244) in the rejection of B16 melanoma cells.
J. Immunol.
174
:
800
807
.
26
Malarkannan
S.
,
Regunathan
J.
,
Chu
H.
,
Kutlesa
S.
,
Chen
Y.
,
Zeng
H.
,
Wen
R.
,
Wang
D.
.
2007
.
Bcl10 plays a divergent role in NK cell-mediated cytotoxicity and cytokine generation.
J. Immunol.
179
:
3752
3762
.
27
Manoussaka
M.
,
Georgiou
A.
,
Rossiter
B.
,
Shrestha
S.
,
Toomey
J. A.
,
Sivakumar
P. V.
,
Bennett
M.
,
Kumar
V.
,
Brooks
C. G.
.
1997
.
Phenotypic and functional characterization of long-lived NK cell lines of different maturational status obtained from mouse fetal liver.
J. Immunol.
158
:
112
119
.
28
Kumar
V.
,
McNerney
M. E.
.
2005
.
A new self: MHC-class-I-independent natural-killer-cell self-tolerance.
Nat. Rev. Immunol.
5
:
363
374
.
29
McNerney
M. E.
,
Guzior
D.
,
Kumar
V.
.
2005
.
2B4 (CD244)–CD48 interactions provide a novel MHC class I-independent system for NK cell self-tolerance in mice.
Blood
106
:
1337
1340
.
30
Sivori
S.
,
Falco
M.
,
Marcenaro
E.
,
Parolini
S.
,
Biassoni
R.
,
Bottino
C.
,
Moretta
L.
,
Moretta
A.
.
2002
.
Early expression of triggering receptors and regulatory role of 2B4 in human natural killer cell precursors undergoing in vitro differentiation.
Proc. Natl. Acad. Sci. USA
99
:
4526
4531
.
31
Kiel
M. J.
,
Yilmaz
O. H.
,
Iwashita
T.
,
Yilmaz
O. H.
,
Terhorst
C.
,
Morrison
S. J.
.
2005
.
SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells.
Cell
121
:
1109
1121
.
32
Holmberg
L. A.
,
Miller
B. A.
,
Ault
K. A.
.
1984
.
The effect of natural killer cells on the development of syngeneic hematopoietic progenitors.
J. Immunol.
133
:
2933
2939
.
33
Desponts
C.
,
Hazen
A. L.
,
Paraiso
K. H.
,
Kerr
W. G.
.
2006
.
SHIP deficiency enhances HSC proliferation and survival but compromises homing and repopulation.
Blood
107
:
4338
4345
.
34
Ghansah
T.
,
Paraiso
K. H.
,
Highfill
S.
,
Desponts
C.
,
May
S.
,
McIntosh
J. K.
,
Wang
J. W.
,
Ninos
J.
,
Brayer
J.
,
Cheng
F.
, et al
.
2004
.
Expansion of myeloid suppressor cells in SHIP-deficient mice represses allogeneic T cell responses.
J. Immunol.
173
:
7324
7330
.
35
Trotta
R.
,
Parihar
R.
,
Yu
J.
,
Becknell
B.
,
Allard
J.
 2nd
,
Wen
J.
,
Ding
W.
,
Mao
H.
,
Tridandapani
S.
,
Carson
W. E.
,
Caligiuri
M. A.
.
2005
.
Differential expression of SHIP1 in CD56bright and CD56dim NK cells provides a molecular basis for distinct functional responses to monokine costimulation.
Blood
105
:
3011
3018
.
36
Tangye
S. G.
,
Lazetic
S.
,
Woollatt
E.
,
Sutherland
G. R.
,
Lanier
L. L.
,
Phillips
J. H.
.
1999
.
Cutting edge: human 2B4, an activating NK cell receptor, recruits the protein tyrosine phosphatase SHP-2 and the adaptor signaling protein SAP.
J. Immunol.
162
:
6981
6985
.
37
Roncagalli
R.
,
Taylor
J. E.
,
Zhang
S.
,
Shi
X.
,
Chen
R.
,
Cruz-Munoz
M. E.
,
Yin
L.
,
Latour
S.
,
Veillette
A.
.
2005
.
Negative regulation of natural killer cell function by EAT-2, a SAP-related adaptor.
Nat. Immunol.
6
:
1002
1010
.
38
Ott
V. L.
,
Tamir
I.
,
Niki
M.
,
Pandolfi
P. P.
,
Cambier
J. C.
.
2002
.
Downstream of kinase, p62(dok), is a mediator of Fc gamma IIB inhibition of Fc epsilon RI signaling.
J. Immunol.
168
:
4430
4439
.
39
Spaggiari
G. M.
,
Carosio
R.
,
Pende
D.
,
Marcenaro
S.
,
Rivera
P.
,
Zocchi
M. R.
,
Moretta
L.
,
Poggi
A.
.
2001
.
NK cell-mediated lysis of autologous antigen-presenting cells is triggered by the engagement of the phosphatidylinositol 3-kinase upon ligation of the natural cytotoxicity receptors NKp30 and NKp46.
Eur. J. Immunol.
31
:
1656
1665
.
40
Held
W.
,
Raulet
D. H.
.
1997
.
Expression of the Ly49A gene in murine natural killer cell clones is predominantly but not exclusively mono-allelic.
Eur. J. Immunol.
27
:
2876
2884
.
41
Johansson
M. H.
,
Bieberich
C.
,
Jay
G.
,
Kärre
K.
,
Höglund
P.
.
1997
.
Natural killer cell tolerance in mice with mosaic expression of major histocompatibility complex class I transgene.
J. Exp. Med.
186
:
353
364
.
42
Chuang
S. S.
,
Lee
J. K.
,
Mathew
P. A.
.
2003
.
Protein kinase C is involved in 2B4 (CD244)-mediated cytotoxicity and AP-1 activation in natural killer cells.
Immunology
109
:
432
439
.
43
Mathew
S. O.
,
Vaidya
S. V.
,
Kim
J. R.
,
Mathew
P. A.
.
2007
.
Human natural killer cell receptor 2B4 (CD244) down-regulates its own expression by reduced promoter activity at an Ets element.
Biochem. Biophys. Res. Commun.
355
:
483
487
.
44
Held
W.
,
Kunz
B.
,
Lowin-Kropf
B.
,
van de Wetering
M.
,
Clevers
H.
.
1999
.
Clonal acquisition of the Ly49A NK cell receptor is dependent on the trans-acting factor TCF-1.
Immunity
11
:
433
442
.
45
Marusina
A. I.
,
Kim
D. K.
,
Lieto
L. D.
,
Borrego
F.
,
Coligan
J. E.
.
2005
.
GATA-3 is an important transcription factor for regulating human NKG2A gene expression.
J. Immunol.
174
:
2152
2159
.
46
Kalesnikoff
J.
,
Baur
N.
,
Leitges
M.
,
Hughes
M. R.
,
Damen
J. E.
,
Huber
M.
,
Krystal
G.
.
2002
.
SHIP negatively regulates IgE + antigen-induced IL-6 production in mast cells by inhibiting NF-kappa B activity.
J. Immunol.
168
:
4737
4746
.
47
Tassi
I.
,
Cella
M.
,
Gilfillan
S.
,
Turnbull
I.
,
Diacovo
T. G.
,
Penninger
J. M.
,
Colonna
M.
.
2007
.
p110gamma and p110delta phosphoinositide 3-kinase signaling pathways synergize to control development and functions of murine NK cells.
Immunity
27
:
214
227
.
48
Tiwari
S.
,
Choi
H. P.
,
Matsuzawa
T.
,
Pypaert
M.
,
MacMicking
J. D.
.
2009
.
Targeting of the GTPase Irgm1 to the phagosomal membrane via PtdIns(3,4)P(2) and PtdIns(3,4,5)P(3) promotes immunity to mycobacteria.
Nat. Immunol.
10
:
907
917
.