Selective induction of OX19+ (CD5+) or OX19- (CD5-) alloreactive cytolytic lymphocytes in the rat

The phenotypes of alloselective cytolytic lymphocytes of the rat are defined by staining of peritoneal cells of alloimmunized donors with monoclonal antibodies, sorting in a cytofluorometer and evaluating cytolytic capacity in a 51Cr-release assay. We demonstrate that alloimmunization of BN rats can result in either OX19+ (CD5+) or OX19- (CD5-) cytolytic alloselective lymphocytes and show that the OX19- (CD5-) cytolytic cells are OX34+ W3/25- (CD4-) OX8+ (CD8+) lymphocytes not exposing surface Ig. It is further demonstrated that the appearance of CD5+ and CD5- cytolytic alloselective lymphocytes are mutually exclusive; immunization with (WF X BN) F1 cells leading exclusively to appearance of OX19+ effector cells while immunization with WF cells leads to OX19- effector cells. Alloimmunization of WF rats only results in appearance of OX19+ cytolytic lymphocytes.     

In vivo induced allo-reactive natural killer cells.

We have previously shown that rat allo-selective cells of the CD2+CD5- phenotype were generated in Brown Norway (BN) rats after immunization with allogeneic Wistar/Furth (WF) cells, whereas immunization with semi-allogeneic F1 (WF/BN) cells generated CD2+CD5+ effector T cells. We now report that the allo-selective CD2+CD5- lymphocytes lacked expression of intact CD3 complexes and expressed NKR-P1 molecules although lower as compared to classical NK cells, implicating that these lymphocytes constitute a subset of NK cells. The CD5+ T cells were not cytolytically active in BN rats immunized with WF cells indicating an intersubset regulation with mutually exclusive activation of either allo-selective T cells or allo-selective NK cells. Cold target inhibition showed that lysis of both allogeneic target cells and NK-sensitive target cells was mediated by the same NKR-P1 intermediate effector cells. These NK cells lysed WF but not allogeneic Fischer 344 or autologous BN target cells, indicating selective recognition of an allogeneic determinant. Semiallogeneic F1 (WF/BN) target cells were not lysed. Furthermore, target cells from F1 (WF/BN) x WF back-cross hybrids lacking expression of RT1n (self-MHC class I) were susceptible to lysis, whereas back-cross hybrids expressing RT1n were protected from lysis, indicating that self-MHC molecules conferred protection from lysis. These findings implicate the existence of NKR-P1intermediate and NKR-P1high NK cell subsets with different regulation and function in vivo.     

Immunogenetics of collagen-induced arthritis in rats. Both MHC and non-MHC gene products determine the epitope specificity of immune response to bovine and chick type II collagens.

Seven inbred, RT1-congenic rat strains were immunized with native bovine (BII), porcine (PII), or chick (CII) type II collagen and observed for onset, incidence, and severity of arthritis. Clinical results were compared with IgG reactive with native rat type II collagen (RII) and the purified, renatured cyanogen-bromide peptides of BII, CII, or RII. Immunodominant responses to CB11, CB9,7, and CB12 of RII were identified. Secondary responses to CB8 and CB10 also occurred. Reproducible patterns of peptide reactivity were defined in each strain and reflected both RT1 and non-RT1 genotypes plus the species of immunizing collagen. BN non-RT1 gene products moderated clinical arthritis but increased the levels of reactivity to CB11 in three strains carrying RT1l,n,av1 haplotypes. WF (RT1u) rats were susceptible to collagen-induced arthritis (CIA) and developed very high levels of autoantibodies with dominant responses to rat CB11 after CII injections and to rat CB11 and CB9,7 after BII injections. DA (RT1av1) rats developed the most severe arthritis but had only moderate (total) levels of anti-RII IgG: a broad response to CB11, CB10, and CB9,7 after CII injections but predominantly to CB12 and CB9,7 after BII injections. Three RT1n strains--DA.1N(BN), WF.1N(MAXX), and BN--were resistant to BII-induced CIA but developed mild arthritis after immunization with CII. After BII: BN IgG reacted with CB9-7, CB11, and CB12; DA.1N and WF.1N IgG reacted with CB9,7 and CB12. After CII: BN IgG reacted broadly with CB11, CB9-7, CB12, and CB8; WF.1N IgG reacted to CB9-7, CB11, CB8, and CB12; DA.1N IgG reacted with CB8, CB11, and CB9-7. Thus, selective induction of CIA in BN, WF.1N, and DA.1N rats by CII correlated with serum IgG reactivity to rat CB11, but overall strain results identified no single cyanogen-bromide peptide as expressing the sole "arthritogenic" epitope in CIA.     

Deficiencies in gut NK cell number and function precede diabetes onset in BB rats

Defects in the intestinal immune system may contribute to the pathogenesis of autoimmune diseases. Intraepithelial lymphocytes represent a substantial fraction of gut-associated lymphocytes, but their function in mucosal immunity is unclear. A newly described population of NK cells that spontaneously secrete IL-4 and IFN-gamma is present in the intraepithelial lymphocyte compartment of the rat. We hypothesized that defects in the number or function of these cells would be present in rats susceptible to autoimmunity. We report that the number of NKR-P1A(+)CD3(-) intraepithelial NK (IENK) cells is deficient before onset of spontaneous autoimmune diabetes in diabetes-prone BB (BBDP) rats. The absolute number of recoverable IENK cells was only approximately 8% of that observed in WF rats. Bone marrow transplantation from histocompatible WF donors reversed the IENK cell deficiency (and prevented diabetes) in these animals, suggesting a hemopoietic origin for their IENK cell defect. Analysis of diabetes-resistant BB rats, which develop autoimmune diabetes only after perturbation of the immune system, revealed IENK cell numbers intermediate between that of BBDP and WF rats. IENK cells were selectively depleted during treatment to induce diabetes. Prediabetic BBDP and diabetes-resistant BB animals also exhibited defective IENK cell function, including decreased NK cell cytotoxicity and reduced secretion of IL-4 and IFN-gamma. IENK functional defects were also observed in LEW and BN rats, which are susceptible to induced autoimmunity, but not in WF, DA, or F344 rats, which are resistant. Defects in IENK cell number and function may contribute to the pathogenesis of autoimmune diseases including type 1 diabetes.     

Interleukin-12 secreted by mature dendritic cells mediates activation of NK cell function

Dendritic cells (DCs) are known to modulate immune response by activating effector cells of both the innate and the adaptive immune system. In the present study, we demonstrate that co-culture of DCs with paraformaldehyde-fixed tumor cells augments the secretion of interleukin (IL)-12 by DCs and these activated DCs upon co-culture with naive NK cells enhance the cytolytic activity of NK cells against NK-sensitive target YAC-1. Similarly, DCs isolated from tumor-bearing animals also activated NK cells in vitro. For efficient activation of NK cells, the ratio of activated DCs to NK cells is crucial. Addition of anti-IL-12 antibody to the culture system completely abolished activation of NK cells by DCs, suggesting that IL-12 secreted by DCs is an essential factor in NK cell activation. Adoptive transfer of DCs isolated from tumor-bearing animals into normal rats also induced activation of NK cells in normal animals.     

NK dendritic cells expanded in IL-15 exhibit antitumor responses in vivo

NK dendritic cells (NKDC) are a novel subtype of DC with NK cell properties. IL-15 is a pleiotropic cytokine that plays an obligate role in the proliferation and survival of NK cells. We hypothesized that IL-15 is also essential for NKDC development. NKDC were nearly absent in IL-15(-/-) mice, but restored by administration of exogenous IL-15. Treatment of wild-type mice with IL-15 caused a 2- to 3-fold expansion of both NK cells and NKDC. After 7 days of culture with IL-15, sorted splenic NKDC expanded 10-fold while NK cells increased 5-fold. NKDC expanded in IL-15 retained their cytolytic capacity but lost the ability to stimulate naive T cells. Meanwhile, NKDC expanded in IL-15 produced 10 times more IFN-gamma as fresh NKDC and conferred protection in a tumor prevention model. Thus, IL-15 is essential to the proliferation and survival of NKDC and IL-15 expanded NKDC possess antitumor properties.     

Na?ve Rat NK Cells Control the Onset of T Cell Response

NK cell function in the rat is only defined in a rudimentary way due to missing tools for clear NK cell identification. The present study introduces the congenic LEW.BH-NKC rat strain which allows distinct detection of rat NK cells using commercial antibodies. LEW.BH-NKC rats were exposed in vivo to the porcine B cell line L23 by subcutaneous transfer of L23 cell suspension. We used Luciferase transgeneic L23 cells to follow the course of rejection by living imaging. L23 cells were rejected within five days after placement under the skin thus the rejection is mediated by innate immune responses in the first place. Indeed we found increased percentages of NK cells in the blood, spleen and in draining lymph nodes using flow cytometry methods. Surprisingly, we found as a consequence a decrease in proliferative T cell response in the draining lymph nodes. We identified NK cells as mediators of this regulation by in vitro performed mixed lymphocyte reactions. The remarkable feature was the naive state of NK cells exhibiting the regulative capacity. Furthermore, the regulation was not exclusively mediated by IL-10 as it has been reported before for influence of T cell response by activated NK cells but predominantly by TGF-β. Interestingly, after initiation of the adaptive immune response, NK cells failed to take influence on the proliferation of T cells. We conclude that naive NK cells build up a threshold of activation impulse that T cells have to overcome.     

Antibody-dependent cell lysis by NK cells is preserved after sarcoma-induced inhibition of NK cell cytotoxicity

Osteosarcoma and Ewing’s sarcoma tumor cells are susceptible to IL15-induced or antibody-mediated cytolytic activity of NK cells in short-term cytotoxicity assays. When encountering the tumor environment in vivo, NK cells may be in contact with tumor cells for a prolonged time period. We explored whether a prolonged interaction with sarcoma cells can modulate the activation and cytotoxic activity of NK cells. The 40 h coculture of NK cells with sarcoma cells reversibly interfered with the IL15-induced expression of NKG2D, DNAM-1 and NKp30 and inhibited the cytolytic activity of NK cells. The inhibitory effects on receptor expression required physical contact between NK cells and sarcoma cells and were independent of TGF-β. Five days pre-incubation of NK cells with IL15 prevented the down-regulation of NKG2D and cytolytic activity in subsequent cocultures with sarcoma cells. NK cell FcγRIIIa/CD16 receptor expression and antibody-mediated cytotoxicity were not affected after the coculture. Inhibition of NK cell cytotoxicity was directly linked to the down-regulation of the respective NK cell-activating receptors. Our data demonstrate that the inhibitory effects of sarcoma cells on the cytolytic activity of NK cells do not affect the antibody-dependent cytotoxicity and can be prevented by pre-activation of NK cells with IL15. Thus, the combination of cytokine-activated NK cells and monoclonal antibody therapy may be required to improve tumor targeting and NK cell functionality in the tumor environment.     

Expression of a laminin-like substance on the surface of murine natural killer (NK) lymphocytes and its role in NK recognition of tumor target cells

We have identified a structure on the surface of murine NK cells that is immunochemically cross-reactive with laminin. Treatment of normal CBA/J spleen cells with monospecific anti-laminin serum plus complement completely eliminated NK cytolytic activity against YAC-1 or RL male 1 target cells. In the absence of added complement, spleen cells preincubated with anti-laminin serum were also reduced in their cytolytic activity due to a reduced capacity to bind to the target cells. Treatment with anti-asialo GM1 serum plus complement also eliminated NK activity, but pretreatment of NK cells with anti-asialo GM1 in the absence of complement did not reduce cytolytic activity. Thus, anti-laminin and anti-asialo GM1 bind to structures on the surface of NK cells that distinguish functional (laminin) from nonfunctional (asialo GM1) sites. Flow cytometric analysis revealed that approximately 15% of normal nonadherent splenic lymphocytes expressed laminin-like structures, whereas 16% expressed asialo GM1 and 19% expressed the NK alloantigen NK 2.1. Treatment of alloimmune cytotoxic T lymphocytes (CTL) with anti-laminin plus complement did not affect CTL activity. Thus, anti-laminin serum appears to detect a cell surface structure present on the NK subset of lymphocytes.     

Chronic shift-lag alters the circadian clock of NK cells and promotes lung cancer growth in rats

Prolonged subjection to unstable work or lighting schedules, particularly in rotating shift-workers, is associated with an increased risk of immune-related diseases, including several cancers. Consequences of chronic circadian disruption may also extend to the innate immune system to promote cancer growth, as NK cell function is modulated by circadian mechanisms and plays a key role in lysis of tumor cells. To determine if NK cell function is disrupted by a model of human shift-work and jet-lag, Fischer (344) rats were exposed to either a standard 12:12 light-dark cycle or a chronic shift-lag paradigm consisting of 10 repeated 6-h photic advances occurring every 2 d, followed by 5-7 d of constant darkness. This model resulted in considerable circadian disruption, as assessed by circadian running-wheel activity. NK cells were enriched from control and shifted animals, and gene, protein, and cytolytic activity assays were performed. Chronic shift-lag altered the circadian expression of clock genes, Per2 and Bmal1, and cytolytic factors, perforin and granzyme B, as well as the cytokine, IFN-γ. These alterations were correlated with suppressed circadian expression of NK cytolytic activity. Further, chronic shift-lag attenuated NK cell cytolytic activity under stimulated in vivo conditions, and promoted lung tumor growth following i.v. injection of MADB106 tumor cells. Together, these findings suggest chronic circadian disruption promotes tumor growth by altering the circadian rhythms of NK cell function.     

Dendritic cells mediate NK cell help for Th1 and CTL responses: two-signal requirement for the induction of NK cell helper function

Early stages of viral infections are associated with local recruitment and activation of dendritic cells (DC) and NK cells. Although activated DC and NK cells are known to support each other's functions, it is less clear whether their local interaction in infected tissues can modulate the subsequent ability of migrating DC to induce T cell responses in draining lymph nodes. In this study, we report that NK cells are capable of inducing stable type 1-polarized "effector/memory" DC (DC1) that act as carriers of NK cell-derived helper signals for the development of type 1 immune responses. NK cell-induced DC1 show a strongly elevated ability to produce IL-12p70 after subsequent CD40 ligand stimulation. NK-induced DC1 prime naive CD4+ Th cells for high levels of IFN-gamma, but low IL-4 production, and demonstrate a strongly enhanced ability to induce Ag-specific CD8+ T cell responses. Resting NK cells display stringent activation requirements to perform this novel, DC-mediated, "helper" function. Although their interaction with K562 cells results in effective target cell killing, the induction of DC1 requires a second NK cell-activating signal. Such costimulatory signal can be provided by type I IFNs, common mediators of antiviral responses. Therefore, in addition to their cytolytic function, NK cells also have immunoregulatory activity, induced under more stringent conditions. The currently demonstrated helper activity of NK cells may support the development of Th1- and CTL-dominated type 1 immunity against intracellular pathogens and may have implications for cancer immunotherapy.     

Resident peritoneal NK cells

In this study, we describe a new population of NK cells that reside in the normal, uninflamed peritoneal cavity. Phenotypically, they share some similarities with the small population of CD49b(-), CD27(+) immature splenic NK cells, as well as liver NK cells, but they differ in their expression of CD62L, TRAIL, and EOMES. Functionally, the peritoneal NK cells resemble the immature splenic NK cells in their production of IFN-γ, GM-CSF, and TNF-α and in the killing of YAC-1 target cells. We also found that the peritoneum induces different behavior in mature and immature splenic NK cells. When transferred i.v. into RAGγc knockout mice, both populations undergo homeostatic proliferation in the spleen, but only the immature splenic NK cells are able to reach the peritoneum. When transferred directly into the peritoneum, the mature NK cells survive but do not divide, whereas the immature NK cells proliferate profusely. These data suggest that the peritoneum is not only home to a new subset of tissue-resident NK cells, but that it differentially regulates the migration and homeostatic proliferation of immature versus mature NK cells.     

Role of non-RT1 genes in the response of rat lymphocytes to Mycoplasma arthritidis T cell mitogen, concanavalin A and phytohemagglutinin

A comparison of splenic cells from various inbred rat strains indicated that DA, Lewis, Buffalo, August, Wistar Furth, and (LEW X BN)F1 all responded well to the Mycoplasma arthritidis T cell mitogen, phytohemagglutinin and concanavalin A, but cells from BN and MAXX rats were very weakly or nonresponsive. Cells from congenic strains expressing nonresponder background genes, and responder haplotypes at RT1 (BN.1L(LEW), RT1; BN.1A(DA), RT1av1) failed to respond significantly to the mitogens. Rats expressing responder background genes but the nonresponder haplotype at RT1 at RT1 (WF.1N-(MAXX), RT1n) exhibited high responses to all mitogens. The controlling role of non-RT1 genes was confirmed by testing tissue-typed (DA X BN)F2 progeny and (DA X BN)F1 X DA and (DA X BN)F1 X BN progeny. No association was seen between the expression of a/a, a/n, or n/n at RT1 and the degree of response to the mitogens. In contrast, as the proportion of DA non-RT1 genes increased, so did the degree of mitogenic responsiveness. Similar results were obtained by using a partially purified preparation of the mycoplasma T cell mitogen. The results indicated that in the (DA X BN)F1 hybrids, responsiveness to all mitogens was recessive: this contrasts with the (LEW X BN)F1 hybrids in which responsiveness was dominant. Finally, we showed that both responder and nonresponder splenic cells were capable of binding the M. arthritidis mitogen. The data contrast with those obtained with nonresponder mouse strains the cells of which failed to bind mitogen due to the absence of the E alpha chain of the I-E-coded molecule.     

Modulation of perforin, granzyme A, and granzyme B in murine natural killer (NK), IL2 stimulated NK, and lymphokine-activated killer cells by alcohol consumption.

Alcohol consumption in mice suppresses the cytolytic activity of natural killer (NK) and lymphokine-activated killer (LAK) cells through unknown mechanisms. Herein, we found that alcohol consumption decreased target cell-induced release of granzyme A activity in freshly isolated splenic NK cells, in NK cells stimulated for 18 h with 1000 IU/ml of interleukin 2, and in LAK cells. The total activity and protein expression of granzymes A and B also were lower in these cells than in cells isolated from water-drinking mice. Interleukin 2 increased granzyme A protein expression independent of alcohol consumption; however, this increase was associated with decreased enzyme activity. In contrast, granzyme B protein expression and enzymatic activity increased in response to interleukin 2. Perforin activity and protein expression were reduced in LAK cells generated from alcohol-consuming mice. We conclude that the mechanism underlying the suppression of NK and LAK cytolytic activity by alcohol consumption involves the collective reduction of target-induced release, activity, and expression of perforin and granular proteases.     

Two structurally related rat Ly49 receptors with opposing functions (Ly49 stimulatory receptor 5 and Ly49 inhibitory receptor 5) recognize nonclassical MHC class Ib-encoded target ligands

The Ly49 family of lectin-like receptors in rodents includes both stimulatory and inhibitory members. Although NK alloreactivity in mice is regulated primarily by inhibitory Ly49 receptors, in rats activating Ly49 receptors are equally important. Previous studies have suggested that activating rat Ly49 receptors are triggered by polymorphic ligands encoded within the nonclassical class Ib region of the rat MHC, RT1-CE/N/M, while inhibitory Ly49 receptors bind to widely expressed classical class Ia molecules encoded from the RT1-A region. To further investigate rat Ly49-mediated regulation of NK alloreactivity, we report in this study the identification and characterization of two novel paired Ly49 receptors that we have termed Ly49 inhibitory receptor 5 (Ly49i5) and Ly49 stimulatory receptor 5 (Ly49s5). Using a new mAb (mAb Fly5), we showed that Ly49i5 is an inhibitory receptor that recognizes ligands encoded within the class Ib region of the u and l haplotypes, while the structurally related Ly49s5 is an activating receptor that recognizes class Ib ligands of the u haplotype. Ly49s5 is functionally expressed in the high NK-alloresponder PVG strain, but not in the low alloresponder BN strain, in which it is a pseudogene. Ly49s5 is hence not responsible for the striking anti-u NK alloresponse previously described in BN rats (haplotype n), which results from repeated alloimmunizations with u haplotype cells. The present studies support the notion of a complex regulation of rat NK alloreactivity by activating and inhibitory Ly49 members, which may be highly homologous in the extracellular region and bind similar class Ib-encoded target ligands.     

Purine metabolites suppress proliferation of human NK cells through a lineage-specific purine receptor.

NK cell proliferation is suppressed in some patients with cancer by unknown mechanisms. Because purine metabolites released into the extracellular space during cell lysis may affect cell function, we hypothesized that these metabolites could serve as feedback regulators of NK cell proliferation. Sorted NK (CD56+/CD3-) cells were incubated with IL-2 (1000 U/ml) in a 4-day thymidine uptake assay with or without 10-10,000 microM of nucleotides. Adenine nucleotides inhibited NK cell proliferation, with ATP = ADP > 5'-adenylylimidodiphosphate > AMP = adenosine; ADP-ribose and nicotinamide adenine dinucleotide, but not nicotinamide or UTP, caused a dose-dependent suppression of thymidine uptake. A total of 100 microM ATP, a concentration that induced a maximal (80%) inhibition of thymidine uptake, did not inhibit cytotoxic activity against K562 targets. Because NK cells retained the ability to lyse K562 targets 4 days after exposure to 500 microM ATP or 1000 microM adenosine, inhibition of thymidine uptake was not due to cell death. Incubation of NK cells with dibutyryl cAMP and forskolin also suppressed thymidine uptake. Cholera toxin and pertussis toxin suppressed NK cell proliferation. Pertussis toxin did not block the adenine nucleotide effects. Further, ATP, but not adenosine or other nucleotides, markedly increased intracellular cAMP in a dose-dependent manner. The ATP-induced increase in cAMP was specific to cytolytic cells, because CD19+ B cells and CD4+ T cells did not increase their intracellular cAMP. These studies demonstrate that NK proliferation is regulated through purine receptors by adenine nucleotides, which may play a role in decreased NK cell activity. The response to adenine nucleotides is lineage-specific.     

Fractalkine-mediated endothelial cell injury by NK cells

Endothelial cells (ECs) are primary targets of immunological attack, and their injury can lead to vasculopathy and organ dysfunction in vascular leak syndrome and in rejection of allografts or xenografts. A newly identified CX3C-chemokine, fractalkine, expressed on activated ECs plays an important role in leukocyte adhesion and migration. In this study we examined the functional roles of fractalkine on NK cell activity and NK cell-mediated endothelial cell injury. Freshly separated NK cells expressed the fractalkine receptor (CX3CR1) determined by FACS analysis and efficiently adhered to immobilized full-length fractalkine, but not to the truncated forms of the chemokine domain or mucin domain, suggesting that fractalkine functions as an adhesion molecule on the interaction between NK cells and ECs. Soluble fractalkine enhanced NK cell cytolytic activity against K562 target cells in a dose- and time-dependent manner. This enhancement correlated well with increased granular exocytosis from NK cells, which was completely inhibited by the G protein inhibitor, pertussis toxin. Transfection of fractalkine cDNA into ECV304 cells or HUVECs resulted in increased adhesion of NK cells and susceptibility to NK cell-mediated cytolysis compared with control transfection. Moreover, both enhanced adhesion and susceptibility of fractalkine-transfected cells were markedly suppressed by soluble fractalkine or anti-CX3CR1 Ab. Our results suggest that fractalkine plays an important role not only in the binding of NK cells to endothelial cells, but also in NK cell-mediated endothelium damage, which may result in vascular injury.     

Spatial organization of signal transduction molecules in the NK cell immune synapses during MHC class I-regulated noncytolytic and cytolytic interactions

The cytolytic activity of NK cells is tightly regulated by inhibitory receptors specific for MHC class I Ags. We have investigated the composition of signal transduction molecules in the supramolecular activation clusters in the MHC class I-regulated cytolytic and noncytolytic NK cell immune synapses. KIR2DL3-positive NK clones that are specifically inhibited in their cytotoxicity by HLA-Cw*0304 and polyclonal human NK cells were used for conjugate formation with target cells that are either protected or are susceptible to NK cell-mediated cytotoxicity. Polarization of talin, microtubule-organizing center, and lysosomes occurred only during cytolytic interactions. The NK immune synapses were analyzed by three-dimensional immunofluorescence microscopy, which showed two distinctly different synaptic organizations in NK cells during cytolytic and noncytolytic interactions. The center of a cytolytic synapse with MHC class I-deficient target is comprised of a complex of signaling molecules including Src homology (SH)2-containing protein tyrosine phosphatase-1 (SHP-1). Closely related molecules with overlapping functions, such as the Syk kinases, SYK, and ZAP-70, and adaptor molecules, SH2 domain-containing leukocyte protein of 76 kDa and B cell linker protein, are expressed in activated NK cells and are all recruited to the center of the cytolytic synapse. In contrast, the noncytolytic synapse contains SHP-1, but is lacking other components of the central supramolecular activation cluster. These findings indicate a functional role for SHP-1 in both the cytolytic and noncytolytic interactions. We also demonstrate, in three-cell conjugates, that a single NK cell forms a cytolytic synapse with a susceptible target cell in the presence of both susceptible and nonsusceptible target cells.