β2 integrin induces TCRζ-Syk-phospholipase C-γ phosphorylation and paxillin-dependent granule polarization in human NK cells

Abstract

Cytotoxic lymphocytes kill target cells through polarized release of the content of lytic granules at the immunological synapse. In human NK cells, signals for granule polarization and for degranulation can be uncoupled: Binding of β(2) integrin LFA-1 to ICAM is sufficient to induce polarization but not degranulation, whereas CD16 binding to IgG triggers unpolarized degranulation. In this study, we investigated the basis for this difference. IL-2-expanded human NK cells were stimulated by incubation with plate-bound ligands of LFA-1 (ICAM-1) and CD16 (human IgG). Surprisingly, LFA-1 elicited signals similar to those induced by CD16, including tyrosine phosphorylation of the TCR ζ-chain, tyrosine kinase Syk, and phospholipase C-γ. Whereas CD16 activated Ca(2+) mobilization and LAT phosphorylation, LFA-1 did not, but induced strong Pyk2 and paxillin phosphorylation. LFA-1-dependent granule polarization was blocked by inhibition of Syk, phospholipase C-γ, and protein kinase C, as well as by paxillin knockdown. Therefore, common signals triggered by CD16 and LFA-1 bifurcate to provide independent control of Ca(2+)-dependent degranulation and paxillin-dependent granule polarization.

Figure 1

Tyrosine phosphorylation of TCR ζ and Syk induced by LFA-1 engagement in IL-2 expanded human NK cells. (A) NK cells were stimulated at 37°C for 20 minutes on plates coated with either purified ICAM-1 or BSA. Cells were lysed, and tyrosine phosphorylated proteins were immunoprecipitated with 4G10-agarose. IPs were resolved by 4-20% SDS-PAGE and immunoblotted with biotinylated anti-phosphotyrosine antibody (4G10). The left and right panels represent short and long exposures of the resulting chemiluminescent reaction. (B) NK cells were stimulated and lysed. TCR ζ, DAP12, and the FcR γ chain were immunoprecipitated, resolved by SDS-PAGE, and immunoblotted for phosphotyrosine. The blot was stripped and probed in succession with anti-TCR ζ, anti-DAP12, and anti-FcR γ antibodies. (C) Anti-phosphotyrosine immunoprecipitations prepared as in Fig. 1A were immunoblotted with an anti-Syk antibody.

Figure 2

LFA-1 dependent responses are ablated by inhibition of the Syk kinase. (A) NK cells pretreated for 45 minutes at 37°C in culture medium with 1 μM Syk inhibitor II or DMSO carrier were mixed with Celltracker Green labeled S2 cells or S2 cells expressing ICAM-1 for 20 min at 37°C, adhered to poly-D-lysine coated slides, fixed, permeablized, and stained with an anti-perforin antibody and Alexa-568 goat-anti-mouse secondary antibody. A representative image of treated or untreated NK cells in contact with S2-ICAM-1 cells is shown. (B) Treated or untreated NK cells in conjugate with S2 or S2-ICAM-1 cells were scored for polarization of perforin toward the target cell interface. Error bars represent S.D. of three experiments conducted with NK cells from different donors. *p < 0.05 relative to DMSO-treated NK cells + S2-ICAM-1 cells. (C) NK cells were pretreated for 45 minutes at 37°C in culture medium with 1μM Syk inhibitor II (Calbiochem) or DMSO carrier. Cells were then stimulated with purified ICAM-1 on plates for 20 minutes at 37°C and lysed. Pyk2, CasL, and paxillin immunoprecipitations and whole cell lysates (WCL) were resolved by SDS-PAGE and immunoblotted for phosphotyrosine.

Figure 3

LFA-1 and CD16 both induce phosphorylation of Syk and TCR ζ, but provoke different cellular responses. (A) NK cells were stimulated with purified ICAM-1 or purified human IgG on plates for either 5 minutes or 20 minutes at 37°C and lysed. Lysates were immunoprecipitated with 4G10 agarose, and immunoprecipitates were resolved by SDS-PAGE and immunoblotted for phosphotyrosine. (B) 4G10-agarose immunoprecipitates from NK cells stimulated as in Fig. 3C for 20 minutes were immunoblotted for Syk and TCR ζ. (C) NK cells were mixed with S2 cells, S2 cells precoated with a polyclonal antiserum (S2 + ADCC), or S2-ICAM-1 cells in the presence of FITC-labeled anti-CD107a antibody and 6 μM monensin for 2 hours at 37°C. Cells were then stained with a PECy7-labeled anti-CD56 antibody at 4°C for 30 minutes. Cells were analyzed by flow cytometry. CD56⁺ NK cells were gated and examined for CD107a staining. (D) NK cells were labeled with Fluo-4 and Fura Red and stimulated (as described in Methods) with either antibodies against CD56, CD16 and CD11a (left panel) or S2 cells, S2 cells + antiserum, or S2-ICAM-1 cells (right panel). Fluo-4 and Fura Red intensities were monitored for 5 minutes total, the ratio of Fluo-4 to Fura Red was calculated, and the ratios for each time course were normalized to the starting ratio at Time=0. Data are expressed as fold changes in the ratio from the starting time point.

Figure 4

PLC-γ and PKC are required for perforin polarization in response to LFA-1 engagement. (A) NK cells were stimulated at 37°C for 20 minutes on plates coated with BSA, purified ICAM-1, or purified human IgG. Cells were lysed, and tyrosine phosphorylated proteins were immunoprecipitated with 4G10-agarose. IPs were resolved by 4-20% SDS-PAGE and immunoblotted for PLC-γ1 and PLC-γ2. (B) NK cells were pretreated in culture medium for 45 minutes with 50 nM PLC inhibitor (U73122) or its inactive analog (U73343). Perforin polarization was analyzed as in Fig. 2B. Error bars represent S.D. of three experiments conducted with NK cells from different donors. *p < 0.05 relative to control U73343-treated NK cells + S2-ICAM-1 cells. (C) Whole cell lysates from cells stimulated as in Fig. 4A were resolved by SDS-PAGE and immunoblotted with an anti-PKC substrate (anti-phospho-serine motif) antibody. (D) NK cells were pretreated in culture medium for 45 minutes with 5 μM bisindolylmaliemide or DMSO carrier at 37°C. Perforin polarization was analyzed as in Fig. 2B. Error bars represent S.D. of three experiments conducted with NK cells from different donors. *p < 0.05 relative to control DMSO-treated NK cells + S2-ICAM-1 cells.

Figure 5

Paxillin and LAT are differentially involved in the LFA-1 and CD16 responses. (A) Anti-phosphotyrosine immunoprecipitates from NK cells stimulated as in Fig. 4A were resolved by SDS-PAGE and immunoblotted for paxillin or LAT. (B) NK cells were nucleofected with 300 pmols of siRNA targeting either paxillin or LAT, or a control siRNA. 48 hours after nucleofection, whole cell lysates from 2.5 × 10⁵ cells were resolved by SDS-PAGE and immunoblotted for paxillin or LAT. (C) NK cells transfected with control, paxillin, or LAT siRNA mixed with either S2 cells or S2 cells precoated with a polyclonal antiserum (S2 + ADCC) were analyzed for CD107a surface expression (degranulation) as in Fig. 3A. Error bars represent S.D. of three experiments conducted with NK cells from different donors. *p < 0.05 relative to control siRNA transfected NK cells + S2+ADCC. (D) NK cells transfected with control, paxillin, or LAT siRNA mixed with either S2 or S2-ICAM-1 cells were analyzed for perforin polarization as in Fig. 2B. Error bars represent S.D. of three experiments conducted with NK cells from different donors. *p < 0.05 relative to control siRNA transfected NK cells + S2-ICAM-1 cells.