NK cytolytic activity is dependent on LFA-1. (A) Fresh lymphocytes were tested for natural cytolysis of CMFDA-labeled K562 or Daudi tumor cell lines in a 4-hour flow cytometric killing assay at 1:12.5 target cell/NK cell target ratio. Samples were incubated with α-NKG2D, α-KIR, or α–LFA-1 (clone TS1/22) antibodies (10 μg/mL). (B) Lymphocytes were treated with IL-2 or IL-12 (200 ng/mL) for 12 hours and incubated with K562 targets at the indicated target/NK ratio and specific lysis was assessed by flow cytometry. (C) Lymphocytes were treated with IL-2 (200 ng/mL) or ICAM-2 (1 μg) and incubated with CMDFA-labeled K562 cells. Lymphocytes were prestained with CD56-PerCPCy5.5 for 10 minutes, mixed with K562 cells at 1:12.5 target cell/NK ratio for 10 minutes, and fixed directly in 2% PFA. Conjugates were assessed as previously described.20 (D) PBMCs depleted for adherent cells were stained for surface CD3-PerCPCy5.5, CD8-PE, CD56-APC, and intracellular perforin-FITC expression. Quadrant gating for CD3 and CD8 is color coded and displayed on CD56 and perforin parameters for gated populations. Distinction between CD56bright and CD56^dim and the distinctions among perforin low, medium, and high levels are displayed. Subset frequencies for intracellular perforin levels in CD3^+/–CD8^+/–CD56^+/– subsets are displayed in Table 1. Numbers in each quadrant correspond to population frequencies within those quadrants. (E) ICAM binding to CD56⁺CD8⁺ NK cells and CD3⁺CD8⁺ cells. The 1 × 10⁶ cells that were cultured in the presence or absence of IL-2 (100 ng/mL, 12 hours) were labeled with ICAM-1, ICAM-2, or ICAM-3 fusion proteins (1.25 μg) at 37°C for 15 minutes in the presence of mouse IgG MOPC21 (4 μg/mL) and human sera. Cells were washed and stained with goat antihuman FC-AX488 (Fab′)₂, and CD56-PE, CD8-PerCPCy5.5, and CD3-APC. Geometric means for ICAM-AX488 are displayed for CD3^– CD56⁺CD8⁺ and CD3⁺CD8⁺ gated populations. Error bars in panels A, C, and E indicate standard deviation of at least 3 experiments.