|Home | About | Journals | Submit | Contact Us | Français|
S.H.L.: conception and design, collection and/or assembly of data, manuscript writing; S.Y., M.J.K. and J.L.: collection and/or assembly of data; J.W.C., T.D.K., S.R.Y., and P.D.G.: data analysis and interpretation; I.C.: study conception and design, financial support, manuscript writing.
Cross-linking of NK activating receptors activates PLC-γ and subsequently induces DAG and Ca2+ as second messengers of signal transduction. Previous studies reported that Ras guanyl nucleotide-releasing protein (RasGRP) 1, which is activated by DAG and Ca2+, is crucial for T cell receptor-mediated Ras-ERK activation. We now report that RasGRP1, which can also be detected in human NK cells, plays an essential role in NK cell effector functions. To examine the role of RasGRP1 in NK cell functions, the expression of RasGRP1 was suppressed using RNAi. Knockdown of RasGRP1 significantly blocked ITAM-dependent cytokine production as well as NK cytotoxicity. Biochemically, RasGRP1-knockdown NK cells showed markedly decreased ability to activate Ras, ERK and JNK. Activation of the Ras-MAPK pathway was independently shown to be indispensable for NK cell effector functions via the use of specific pharmacological inhibitors. Our results reveal that RasGRP1 is required for the activation of the Ras-MAPK pathway leading to NK cell effector functions. Moreover, our data suggest that RasGRP1 might act as an important bridge between PLC-γ activation and NK cell effector functions via the Ras-MAPK pathway.
Natural killer (NK) cells are effector lymphocytes that eliminate tumor cells or virus-infected cells (1). NK cells also have the ability to produce immunoregulatory cytokines, such as interferon-γ (IFN-γ), granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor-α (TNF-α) (2). These NK cell effector functions are regulated by multiple activating and inhibitory NK cell receptors (3, 4). Whereas inhibitory NK cell receptors transmit inhibitory signals to the downstream signaling molecules through their cytoplasmic ITIM, activating NK cell receptors signal through ITAM-containing adaptors, such as CD3ζ, FcRγ, and DAP12, or the cytoplasmic Tyr-Ile-Asn-Met (YINM) motif-containing adaptor DAP10 (5). Upon engagement of activating receptors, such as natural cytotoxicity receptors (NCRs) and NKG2D, tyrosine phosphorylation in the ITAM or YINM motif of adaptor molecules is mediated by Src family kinases, leading to the subsequent activation of a variety of downstream signaling molecules, including the Vav family, PI3K, and phospholipase C-γ (PLC-γ). Finally, transmitted signals result in activation of mitogen-activated protein kinases (MAPKs), which are crucial for cytolytic granule release and cytokine generation (6, 7). Among these molecules propagating activation signals, PLC-γ plays an essential role in the signal transduction leading to NK cell cytotoxicity and cytokine production (8-10). PLC-γ, which hydrolyzes membrane phosphatidylinositols into inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG) (11, 12), modulates the intracellular calcium ion concentration (Ca2+) and activates downstream-signaling cascades for NK cell effector functions.
Ras guanyl nucleotide-releasing protein 1 (RasGRP1) is a member of the RasGRP family that contain a DAG-binding C1 domain, a Ras exchange motif, calcium-binding EF-hands, and a guanine-nucleotide exchange factor (GEF) domain (13). Following PLC-γ activation, RasGRP1 is activated by DAG and Ca2+ leading to activation of the Ras family via guanine-nucleotide exchange with dissociation of GDP from Ras and association with GTP (14, 15). RasGRP1 plays an essential role in T cell receptor (TCR) signaling. Thymocytes isolated from RasGRP1−/− mice display a defect in Ras-mediated extracellular signal-regulated protein kinase (ERK) activation in response to TCR stimulation, and a block in the double-positive stage of thymocyte development (16). In addition, RasGRP1 plays an important role in IgE-mediated signal transduction and mast cell function. IgE-mediated degranulation is impaired in mast cells of RasGRP1−/− mice, and these mice display severely defective IgE-evoked systemic anaphylaxis (17).
Despite the evidence that RasGRP1 plays a critical role in TCR-mediated and IgE-mediated signaling, whether it functions in NK cell receptor signaling has not been examined. Because activation of Ras has been shown to be involved in NK cell-mediated cytotoxicity and IFN-γ production (6, 18, 19), we examined if RasGRP1 plays a role in NK cell activation and identified an essential role in effector functions and receptor-mediated Ras-MAPK activation.
Human primary NK cells were isolated from the umbilical cord blood (UCB) using the human NK Cell Isolation Kit (Miltenyi). The isolated primary NK cell populations which were >95% CD56+CD3− were cultured in Myelocult H5100 (StemCell Technologies) supplemented with IL-15 (10 ng/ml). The UCB was collected from umbilical veins after neonatal delivery, with informed consent from the pregnant mothers and following the guidance of the local institutional review board (IRB). To prepare hematopoietic stem cell (HSC)-derived mature NK cells, CD34+ HSCs were isolated from UCB using the CD34 MicroBead Kit (Miltenyi). CD34+ HSCs were differentiated into NK cell precursors by incubating the cells in Myelocult H5100 supplemented with SCF (30 ng/ml) and Flt3 ligand (50 ng/ml) for 14 days. NK cell precursors were differentiated into mature NK cells by stimulation with IL-15 (30 ng/ml) for an additional 14 days. Mature NK cells (>95% CD56+CD3− cells) were maintained in Myelocult H5100 with 10 ng/ml IL-15 and used for the functional assays. The human NK cell line NK-92 (American Type Culture Collection, ATCC) was cultured in alpha MEM (Gibco), supplemented with 20% heat-inactivated FBS (Hyclone), 2 mM l-glutamine and IL-15 (10 ng/ml). The K562 cell line (ATCC) was cultured in RPMI 1640 (Gibco) supplemented with 10% heat-inactivated FBS (Hyclone). Recombinant human SCF, Flt3 ligand, IL-12, and IL-15 were purchased from PeproTech. FTI-277, PD98059, SP600125 and SB203580 were purchased from Calbiochem.
For lentiviral infection of NK-92 cells, the pLKO.1-non-target shRNA control vector (SHC002) and pLKO.1-RasGRP1 shRNA vector (TRCN0000048268) were purchased from Sigma. Lentiviruses were produced using a third-generation packaging system (pMDLg/pRRE, pRSV-Rev and pMD2.G) in HEK293T cells. The lentivirus-containing supernatants were cleared by centrifugation at 3000 rpm for 5 min at 4 °C, passed through a 0.45 μm filter, and concentrated by ultracentrifugation at 50,000 g for 90 min at 4 °C. Upon infection, shRNA-expressing NK-92 cells were selected with puromycin (2 μg/ml) for 3 weeks before functional analysis. Nucleofections of NK cell precursors and mature NK cells were performed using the Amaxa Human CD34 Cell Nucleofector™ Kit (program U-08), and the non-target control siRNA and RasGRP1 siRNA SMARTpool were purchased from Dharmacon.
Cytotoxicity was examined using a standard 4-h 51Cr-release assay. 51Cr-labeled target K562 cells (3×105 cells/well) and serial dilutions of NK cells were plated in triplicate. The 51Cr released into the supernatant was measured using a γ-counter. The percentage of specific lysis was calculated using the formula: (experimental release - spontaneous release)/(maximum release - spontaneous release)×100. To evaluate cytokine secretion, NK cells were stimulated in duplicate for 16 h with plate-bound antibodies (10 μg/ml), IL-12 (20 ng/ml), or PMA/ionomycin (1 or 2 ng/ml, 0.1 or 0.2 μg/ml). The secretion of IFN-γ (Assay Designs), TNF-α, and GM-CSF (R&D Systems) into the supernatant was measured by ELISA.
Total RNA was extracted using TRIZOL (Invitrogen) and reverse transcribed into cDNA using M-MLV reverse transcriptase (Promega) with random primers (Takara Bio). Real-time PCR was performed using a Dice TP 800 Thermal Cycler and the SYBR Premix Ex Tag (Takara Bio). The data were normalized to the amount of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) transcript. The primer sequences were: 5′-ggctcaaggagacaagttcg-3′ and 5′-gaagtcggtgcactctccata-3′ for RasGRP1, 5′-gtccaacgcaaagcaataca-3′ and 5′-ctcttcgacctcgaaacagc-3′ for IFN-γ, 5′-gtcctcttcaagggccaag-3′ and 5′-tagtcgggccgattgatct-3′ for TNF-α, 5′-catgatggccagccactac-3′ and 5′-taatctgggttgcacaggaa-3′ for GM-CSF, and 5′-cagcctcaagatcatcagca-3′ and 5′-gtcttctgggtggcagtgat-3′ for GAPDH.
Cells were washed twice with ice-cold PBS and lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.25% SDS, 1% NP-40, and 1 mM EDTA, supplemented with a protease inhibitor cocktail tablet and a phosphatase inhibitor cocktail tablet from Roche). The cell lysates were resolved in 8 or 12% SDS-PAGE gels and transferred to a PVDF membrane (Millipore). The membrane was probed with antibodies specific to the following molecules: p-PLC-γ1Tyr783, p-SrcTyr416, Src, p-ERKThr202,Tyr204, ERK, p-JNKThr183,Tyr185, JNK, p-p38Thr180,Tyr182 and p38 (Cell Signaling); RasGRP1, PLC-γ1 and β-actin (Santa Cruz); and GAPDH (Assay Designs). After incubation with peroxidase-conjugated anti-rabbit or anti-mouse IgG (Jackson ImmunoResearch), the signals were detected using SuperSignal West Pico Chemiluminescent Substrate (Pierce) or Immobilon Western Chemiluminescent HRP Substrate (Millipore).
NK-92 cells (1 × 107 cells) were stimulated with K562 cells (1 × 106 cells) or plate-bound anti-NKp30 mAb (10 μg/ml) for 5 min. Ras activation was assessed with the Ras Activation kit (Assay Designs), according to the manufacturer's instructions. In brief, stimulated NK cells were lysed in the Lysis/Binding/Wash buffer. The lysates were incubated with 80 mg of GST-Raf1-RBD on immobilized glutathione resin for 1 h. The resin was then washed and boiled in 2× SDS sample buffer, and GTP-bound pan-Ras was detected by western blotting with anti-pan-Ras antibody.
Comparisons were analyzed for statistical significance by the Student's t test using Microsoft Excel software. A p value < 0.05 was considered as significant.
To determine the potential role of RasGRP1 in NK cells, we first investigated whether RasGRP1 is expressed in human NK cells. RasGRP1 protein was detected in NK-92 and primary NK cells (Fig. 1A). Additionally, RasGRP1 was expressed in mature NK cells (mNK), but not in NK cell precurors (pNK) (Fig. 1B). Next, we examined whether RasGRP1 functions in NK cytotoxicity. To investigate the cytolytic activity of NK-92 cells in which expression of RasGRP1 was silenced (Fig. 1C), we measured lysis of the NK target, K562 cells (a human erythromyeloblastoid leukemia cell line). RasGRP1 shRNA-expressing NK-92 cells exhibited reduced capacity to kill target K562 cells (Fig. 1D). RasGRP1 siRNA-nucleofected mature NK cells (Fig. 1E) also displayed diminished cytolytic activity compared with control siRNA-treated mature NK cells (Fig. 1F). Thus, RasGRP1 is expressed in human NK cells and functions in NK cytotoxicity, leading us to hypothesize that RasGRP1 might be involved in the K562-evoked NK receptor signaling pathway that leads to expression of cytolytic activity.
After engagement of NK activating receptors that associate with ITAM-containing adaptors, NK cells produce proinflammatory cytokines, such as IFN-γ, TNF-α, and GM-CSF (7, 20, 21), in addition to exhibiting cytotoxicity. To address the impact of RasGRP1 on ITAM-dependent cytokine production, we analyzed NKp30-mediated cytokine generation in control shRNA- and RasGRP1 shRNA-expressing NK-92 cells. NKp30-mediated secretion of IFN-γ, TNF-α, and GM-CSF was significantly reduced in RasGRP1 shRNA-expressing NK-92 cells (Fig. 2A), and transcription of these cytokines was also decreased (Fig. 2B). After cross-linking of NK activating receptors, ITAM-dependent signaling cascades occur including activation of PLC-γ (5, 6). As it has been reported that RasGRP1 is directly activated by PLC-γ-induced DAG and Ca2+ (13, 14, 22), we evaluated the effect of PMA (as a DAG analog) and ionomycin (as a calcium ionophore) on cytokine production in RasGRP1 shRNA-expressing NK-92 cells. PMA/ionomycin-induced cytokine generation was markedly reduced in RasGRP1 shRNA-expressing NK-92 cells compared with control shRNA-expressing NK-92 cells (Fig. 2A, 2B). IFN-γ production in response to IL-12, a major stimulatory cytokine of IFN-γ, was normal in RasGRP1 shRNA-expressing NK-92 cells, but IL-12 did not induce GM-CSF or TNF-α production in NK-92 cells (Fig. 2A, 2B). Similar to RasGRP1 shRNA-expressing NK-92 cells, cytokine production was impaired after stimulation through NCRs (NKp30, NKp44 and NKp46) and by PMA/ionomycin, but not by stimulation of IL-12 receptors in RasGRP1 siRNA-nucleofected mature NK cells (Fig. 2C). Thus, these results indicate that RasGRP1 is essential for ITAM-dependent cytokine production in NK cells.
Our data suggested that impaired intracellular signaling events might be accountable for the defective NK cell effector functions in RasGRP1-knockdown NK cells. In addition, other studies have demonstrated that DAG and Ca2+-bound RasGRP1 play a critical role in the activation of Ras protein (14, 15). Before evaluating the effect of RasGRP1-knockdown on Ras activation, we investigated whether knockdown of RasGRP1 affects activation of PLC-γ, the upstream signaling molecule of RasGRP1. NKp30-mediated activation of PLC-γ1 was normal in RasGRP1 shRNA-expressing NK-92 cells (Fig. 3A). Moreover, knockdown of RasGRP1 did not affect Src phosphorylation, an important kinase for promoting ITAM-mediated proximal signaling events. Next, we examined the Ras activity in RasGRP1-knockdown NK-92 cells using a GST protein fused to the Ras-binding domain (RBD) of Raf1 (GST-Raf1-RBD). The amount of GTP-bound Ras precipitated by GST-Raf1-RBD was reduced in RasGRP1 shRNA-expressing NK-92 cells after incubation with K562 or NKp30 stimulation (Fig. 3B). As shown in Figures Figures11 and and2,2, RasGRP1-knockdown NK cells displayed defective effector functions. To examine if Ras activation is important for NK cell cytotoxicity and cytokine production, we measured NK cell effector functions in NK-92 cells treated with the Ras inhibitor FTI-277. NK cell cytolytic activity was decreased by treatment with FTI-277 in a dose-dependent manner (Fig. 3 C). In addition, NKp30-mediated and PMA/ionomycin-induced production of IFN-γ, TNF-α, and GM-CSF was markedly reduced in FTI-277-treated NK-92 cells (Fig. 3D). However, treatment with FTI-277 did not affect IL-12-induced IFN-γ secretion similar to knockdown of RasGRP1. Overall, these data suggest that RasGRP1 regulates Ras activation, and that Ras activation is important for NK cell cytotoxicity and ITAM-dependent cytokine production.
RasGRP1 plays an important role in TCR-mediated Ras-ERK activation, and Ras protein is an essential factor in the regulation of the MAPK pathway (23, 24). Additionally, MAPK activation is crucial for cytolytic granule release and cytokine generation in NK cells (6, 7). Therefore, we investigated whether reduced Ras activity by RasGRP1-knockdown affects the MAPK pathway responsible for NK cell effector functions. To investigate the impact of RasGRP1 on MAPK activation, control shRNA- and RasGRP1 shRNA-expressing NK-92 cells were stimulated with K562, plate-bound anti-NKp30 or PMA/ionomycin. After incubation with K562 target cells, RasGRP1 shRNA-expressing NK-92 cells displayed reduced phosphorylation of ERK and JNK compared with control shRNA-expressing NK-92 cells (Fig. 4A). Similarly, NKp30-mediated and PMA/ionomycin-induced activation of ERK and JNK were decreased in RasGRP1-knockdown NK-92 cells (Fig. 4B, 4C). However, p38 phosphorylation was not reduced in RasGRP1 shRNA-expressing NK-92 cells in response to K562, anti-NKp30 or PMA/ionomycin, indicating that p38 activity is not affected by RasGRP1. Next, we confirmed the role of MAPKs in NK cell effector functions using specific pharmacological inhibitors of MAPKs. Use of the JNK inhibitor SP600125 substantially reduced lysis of K562 target cells, and both the ERK inhibitor PD98059 and the p38 inhibitor SB203580 slightly decreased cytolytic activity (Fig. 4D). Additionally, SP600125 significantly blocked NKp30-mediated IFN-γ production, and PD98059 weakly diminished NKp30-mediated IFN-γ production, but SB203580 did not affect NKp30-mediated IFN-γ production (Fig. 4E). Collectively, these results indicated that RasGRP1 is required for the activation of the ERK and JNK pathways in NK cells, and we conclude that the impairment of Ras-mediated ERK and JNK activation by RasGRP1 RNAi is a probable cause for the defect in cytotoxicity and cytokine production in RasGRP1-knockdown NK cells.
NK cells are large granular lymphocytes of innate immunity that contribute to the killing of tumor or virus-infected cells (1). Additionally, NK cells produce important inflammatory cytokines, such as IFN-γ, TNF-α, and GM-CSF (2). These effector functions of NK cells are controlled by the coordinated balance between activating and inhibitory receptors (4). The inhibitory NK receptors suppress cellular responses via ITIMs present in their cytoplasmic tails. Upon stimulation of the inhibitory receptors, phosphorylated ITIMs bind to the SH2 domains of protein tyrosine phosphatases, such as SHP-1 and SHP-2, which dephosphorylate a number of tyrosine-phosphorylated signaling proteins (25, 26). The activating NK cell receptors associate with adaptor molecules that signal through their cytoplasmic signaling motifs, such as ITAMs, which initiates both cytotoxicity and cytokine production, or the YINM motif, which triggers only cytotoxicity. The NCR family that recognizes unidentified ligands on tumor cells contains three defined members (NKp46, NKp44 and NKp30). NKp44 associates with ITAM-containing DAP12, whereas NKp30 and NKp46 recruit ITAM-bearing CD3ζ and FcRγ as adaptor molecules (5). NKG2D is an activating receptor that recognizes ULBP-1/2/3 (27) and MIC-A/B (28) on the surface of target cells. In mice, NKG2D recruits both ITAM-bearing KARAP/DAP12 and YINM motif-bearing DAP10. In contrast to mice, NKG2D in humans only associates with DAP10. Upon engagement of NK activating receptors including NCRs and NKG2D, tyrosine residues in ITAM or YINM motif are phosphorylated by the recruited Src family kinases (29). Phosphorylated ITAM recruits Syk protein family members, which subsequently activate PLC-γ, whereas the phosphorylated YINM motif directly recruits Grb2-mediated PLC-γ independent of Syk. Activated PLC-γ hydrolyzes membrane phosphatidylinositols into IP3 and DAG for downstream signaling events. PKC-θ has been reported to be a putative signaling protein downstream of PLC-γ in NK cells (30). However, PKC-θ-deficient NK cells displayed a defect only in ITAM-dependent IFN-γ secretion, whereas PLC-γ is required for both NK cytotoxicity and cytokine production (8). In addition, several signaling molecules, such as PDK1, SLP76, and Vav1, are also involved in PKC-θ activation (31-33). Despite the fact that PLC-γ is essential for NK cell effector functions, a major downstream signaling protein of PLC-γ, especially DAG and Ca2+-binding protein, has not been well-defined.
In the present study, we stimulated NK cells with K562 cells, engagement of NCRs, or PMA/ionomycin treatment to induce NK cell effector functions. K562 cells expressed two NKG2D ligands, MICA/B and ULBP2, and treatment with the NKG2D blocking mAb markedly reduced NK cytotoxicity against K562 (34). Thus, the incubation with K562 cells could trigger NKG2D-DAP10-mediated signaling events and cytotoxicity in NK cells. Upon engagement of NCR, the ITAM-mediated signaling pathway is activated, and NK cell effector functions are evoked. Both stimulation with K562 and cross-linking of NCRs activate PLC-γ and produce DAG and IP3-mediated Ca2+ influx (5, 6). PMA is a DAG analog that binds to and activates a DAG target protein. Ionomycin is a calcium ionophore that transports Ca2+ across biological membranes and raises the intracellular Ca2+ level. Therefore, treatment with PMA/ionomycin mimics the induction of DAG and Ca2+, which are produced by receptor-mediated PLC-γ activation, and activate DAG and Ca2+ target signaling proteins that regulate the production of inflammatory cytokines from NK cells (21, 35). In this paper, we demonstrate that RasGRP1 might be a major downstream signaling molecule of PLC-γ in NK cells.
RasGRP1 has the ability to bind to and activate Ras family proteins via a Ras exchange motif, a guanine-nucleotide exchange factor domain (GEF), and is an important regulator in the control of the Ras-mediated signaling pathway. At the beginning of this study, we hypothesized that RasGRP1 plays a role in the NK receptor signaling pathway because it is essential for TCR-mediated and IgE-dependent signaling pathways (16, 17), and it is also expressed in human NK cells. We observed that knockdown of RasGRP1 inhibited NK cell effector functions, including cytotoxicity and ITAM-dependent cytokine production. Similar to RasGRP1-deficient thymocytes, RasGRP1-knockdown NK cells displayed an impaired activation of the Ras-MAPK pathway, suggesting that RasGRP1 connect PLC-γ activation to the Ras-MAPK pathway in NK cells (Fig. 5). MAPKs are known as essential kinases of NK cell effector functions (6, 7). It has also been reported that treatment with the Ras inhibitor FTI-277 significantly reduces ITAM-dependent IFN-γ production in mouse NK cells (19), and a Ras-ERK pathway has been implicated in ADCC (18). However, Wei et al. reported that the spontaneous cytotoxicity is independent of Ras-mediated signaling despite the detection of Ras activation followed by target ligation (36). We treated NK-92 cells with FTI-277 for 30 min before triggering NK cell effector functions, but Wei et al. treated with FTI-277 for 24 h prior to stimulation. It is possible that this difference in pretreatment time might lead to disparate cytotoxicity results. Nevertheless, we confirmed that Ras activation is important for cytotoxicity and cytokine production using a pharmacological specific inhibitor of Ras and knockdown of RasGRP1, a Ras activator. However, further analysis is required to elucidate which Ras member (N-Ras, K-Ras, and/or H-Ras) is activated by RasGRP1 in NK cells.
As shown in Figure 2, knockdown of RasGRP1 did not affect IL-12-induced IFN-γ production. This result was expected because the IL-12 receptor signals mainly through STAT4, and not through Ras proteins (37). Similarly, IL-12-mediated IFN-γ secretion was normal in FTI-277-treated NK-92 cells (Fig. 3). This result also showed that treatment with FTI-277 had targeted activity and did not globally affect cellular activity or survivial of NK-92 cells.
In conclusion, we identified RasGRP1 as an essential regulator of NK cell cytotoxicity and cytokine production. Knockdown of RasGRP1 inhibited activation of Ras, ERK, and JNK, and consequently blocked NK cell effector functions. In addition, our results indicate that RasGRP1 might connect PLC-γ activation to NK cell effector functions via the Ras-MAPK pathway.
1This work was supported in part by grants from the GRL project and the New Drug Target Discovery Project (M10848000352-08N4800-35210), the Ministry of Education, Science & Technology, Republic of Korea.
The authors declare no conflicts of interest.