JNK MAP kinase activation is required for MTOC and granule polarization in NKG2D-mediated NK cell cytotoxicity

Abstract

Interaction of the activating receptor NKG2D with its ligands is a major stimulatory pathway for cytotoxicity of natural killer (NK) cells. Here, the signaling pathway involved after NKG2D ligation is examined. Either incubation of the NKG2D-bearing human NKL tumor cell line with K562 target cells or cross-linking with NKG2D mAb induced strong activation of the mitogen-activated protein (MAP) kinases. Selective inhibition of JNK MAP kinase with four different means of inhibition greatly reduced NKG2D-mediated cytotoxicity toward target cells and furthermore, blocked the movement of the microtubule organizing center (MTOC), granzyme B (a component of cytotoxic granules), and paxillin (a scaffold protein) to the immune synapse. NKG2D-induced activation of JNK kinase was also blocked by inhibitors of Src protein tyrosine kinases and phospholipase PLCgamma, upstream of JNK. Similarly, a second MAP kinase pathway through ERK was previously shown to be required for NK cell cytotoxicity. Thus, activation of two MAP kinase pathways is required for cytotoxic granule and MTOC polarization and for cytotoxicity of human NK cells when NKG2D is ligated.

Fig. 1.

Activation of JNK is required for NKG2D-mediated cytotoxicity. (A) NKL cells were cross-linked with anti-NKG2D mAb for the indicated times at 37°C. (Left) Activation of MAPKs was detected with anti-phospho-JNK, p38, or ERK, and total proteins were measured by blotting with anti-JNK, p38, and ERK. (Right) NKL cells were cultured with K562 cells at a 1:1 ratio for 0–30 min. Activation of JNK was detected by anti-phospho-JNK. Anti-JNK was used as a loading control. (B) Percentage of lysis of target cells by NK cells. NKL (Upper) or pNK (Lower Left) cells, pretreated for 30 min at 37°C with either DMSO, 20 μM p38 MAPK inhibitor SB, 10 μM MEK1/2 inhibitor PD, 10 μM JNK inhibitor SP600125 (SP), or potent peptide inhibitor D-JNK-1 were tested for lysis of ⁵¹Cr-labeled K562 cells. (Lower Right) Murine mastocytoma P815 cells labeled with ⁵¹Cr were coated with anti-NKG2D and then incubated with NKL cells pretreated with JNK inhibitor SP at E:T ratio of 20:1. (C) NKL cells pretreated with 10 μM SP or 20 μM D-JNK-1 were cross-linked with anti-NKG2D mAb for the indicated times. Phosphorylation of JNK was then detected by anti-phospho-JNK antibody. Total JNK or ERK was measured as a loading control. (D) NKL cells pretreated with JNK inhibitor SP at different concentrations were cross-linked with anti-NKG2D mAb for 5 min. Phosphorylation of JNK was then detected by anti-phospho-JNK antibody. Total ERK was measured as a loading control. These bands were run in the same gel as C, Right, and the controls for the two experiments are the same. Data shown in A–D are representative of three or more individual experiments.

Fig. 2.

Inhibition of JNK activity by a JNK dominant-negative mutant or depletion of JNK by RNAi reduces NK killing activity. (A) NKL cells were transiently transfected with the GFP-JNKAF (JNKdn) mutant by using the Amaxa nucleofection system. Expression of GFP-JNKdn was confirmed by FACS analysis (Left) and Western blot analysis with anti-JNK antibody (Right). The lower band is endogenous JNK, and the upper band is the fusion protein GFP-JNKdn. (B) NKL cells expressing JNKdn together with NKL cells expressing GFP or GFP-p38AF (p38dn) as control were tested for lysis of ⁵¹Cr-labeled K562 cells (E:T = 20:1). (C) NKL cells were transfected with siRNAs specific for JNK-1, p38, or negative control siRNA. Depletion of JNK-1 expression was confirmed by Western blotting with anti-JNK antibody (Upper) or with anti-p38 as a control (Lower). (D) NKL cells treated with JNK-1, p38, or negative control siRNAs were tested for lysis of ⁵¹Cr-labeled K562 cells (E:T = 20:1, Right). Data are representative of three or more experiments.

Fig. 3.

Effect of specific inhibition of JNK on synapse formation between NK cells and K562 cells. (A) Fixed pNK-K562 conjugates were stained with rabbit polyclonal anti-JNK that also reacts with phospho-JNK or anti-phospho-JNK, followed by the staining with Alexa Fluor 568-conjugated goat anti-rabbit IgG. (B) NKL cells pretreated with 10 μM D-JNK-1 for 30–45 min at 37°C were incubated with K562 cells for 30 min, allowing formation of cell–cell conjugates. After cell fixation and permeabilization, microtubules (MT) were stained with anti-α-tubulin mAb, followed by Alexa Fluor 568-conjugated anti-mouse IgG (red). F-actin (green) was stained with phalloidin-Alexa Fluor 488, and granules (blue) were stained with Alexa Fluor 647-conjugated anti-granzyme B mAb. Representative images are shown. Average percentages of NKL conjugates with MTOC or granule polarization are shown with SD and P values after counting >50 conjugates. Data are representative of three or more independent experiments. See Materials and Methods and SI Fig. 8B for details of determination of polarization.

Fig. 4.

JNK and paxillin. (A) pNK cells treated with 10 μM SP600125 (SP) or without (C, control) JNK inhibitor were incubated with K562 cells and then stained with mouse anti-actin (green) and monoclonal anti-paxillin (red) in both NK cells (upper cell) and target cell (lower cell). (B) Percentage of paxillin polarization to the synapse of pNK cells treated or not treated with JNK inhibitor. For each effector–target cell, 40 conjugates were examined. (C) NKL cells were transfected with pBABE-GFP-PaxS178A mutant and then tested for the lysis of ⁵¹Cr-labeled K562 cells (E:T = 20:1). (D) GFP-PaxS178A-transfected NKL cells were treated or not treated with JNK inhibitor and then incubated with K562 cells. The conjugates were stained with Alexa Fluor 488-conjugated phalloidin (green) and Alexa Fluor 647-conjugated anti-granzyme B (blue). (E) NKL cells were transfected with paxillin siRNA, GFP siRNA, or control siRNA (mock). (Inset) Depletion of JNK expression was confirmed by Western blotting with anti-paxillin antibody. NKL cells treated with these siRNAs were tested for lysis of ⁵¹Cr-labeled K562 cells. Data are representative of three or more experiments.

Fig. 5.

Upstream regulators for the NKG2D-induced JNK activation. (A) NKL cells were preincubated with 200 nM wortmannin and then incubated with ⁵¹Cr-labeled K562 cells at E:T = 10:1 for 4 h at 37°C. Percentage of lysis was measured as ⁵¹Cr release. (B) NKL cells were pretreated with 200 nM wortmannin and then stimulated by cross-linking with anti-NKG2D monoclonal antibody at 37°C for 0–30 min. NKL were also treated with phorbol 12-myristate 13-acetate (PMA) and NaCl as positive controls. Activation of JNK was detected by anti-phospho-JNK (Upper), and the total protein was measured by anti-JNK (Lower). (C) NKL cells, treated with DMSO, 20 μM Src PTK inhibitor PP2, 100 μM Syk inhibitor piceatannol (Pice), 2 μM PKC inhibitor bisindolylmaleimide (Bis), or 10 μM PLC inhibitor U73122 were incubated with ⁵¹Cr-labeled K562 cells (ratio 20:1) to measure percentage of lysis. (D) NKL cells pretreated with DMSO, PP2, and U73122 were cross-linked with anti-NKG2D mAb for the indicated times. Activation of JNK was detected with anti-phospho-JNK antibody, and total JNK protein was examined by using anti-JNK antibody. Results are representative of three experiments.