Modulation of cortical actin meshwork density restores degranulation of LYST-deficient NK cells. A, Images of F-actin (green) and lytic granules (red) proximal to the contact site formed by LYST CRISPR cells and treated with dimethyl sulfoxide (vehicle), swinholide A, or latrunculin B. Scale bars = 5 μm. Openings in the actin meshwork are shown as heat maps. Close-up images show the synapse area containing lytic granules (white rectangles in the leftmost images). Scale bars = 2 μm. Position of lytic granules relative to the actin filaments is illustrated by Y-Z plane projections and optical sections acquired every 220 nm from the contact site (Z1-Z4). MIP, Maximum intensity projection. B and C, Size of actin openings and penetrability of the actin meshwork at the synapse for cells treated and stimulated as in Fig 5, A. D and E, Granzyme B delivery from NK92mi cells, untransduced (UT) or transduced with EV or LYST CRISPR, and pretreated with dimethyl sulfoxide (vehicle), swinholide A, or latrunculin B. The change in granzyme B substrate fluorescence in target cells was monitored by using flow cytometry (see Fig E5). Fig 5, D, Percentage of target cells positive for granzyme B. Fig 5, E, The intensity of granzyme B substrate fluorescence in target cells. F, Cytotoxicity of NK cells pretreated with dimethyl sulfoxide (—) or lenalidomide (+). The percentage of target cell lysis is shown at a 10:1 and 2:1 effector/target (E:T) ratio for NK92mi and ex vivo NK cells, respectively. Graphs show mean values with SDs determined from 4 to 5 (Fig 5, B-E) or 2 to 3 (Fig 5, F) experiments. *P < .05, **P < .01, ***P < .001, and ****P < .0001, 1-way (Fig 5, B, D, and E) or 2-way (Fig 5, C) ANOVA or paired t test (Fig 5, F).