Supplemental Data Figure S1. Effect of TS2/4 and R6.5 antibodies on the kinetics of CD16.NK-92-mediated specific lysis of SKBR-3 target cells. (A) Specific lysis of IFN-γ-treated SKBR-3 cells in the absence (filled circles) or presence of either non-blocking anti-α L integrin antibody TS2/4 (empty squares) or blocking anti-icam-1 antibody R6.5 (empty diamonds). Cr 51 -labeled SKBR3 cells were combined with CD16.NK-92 cells at 1:5 ratio and incubated with Herceptin antibody (1 µg/ml) in the absence or presence of TS2/4 or R6.5 antibodies at saturated concentrations (5 µg/ml) for indicated periods of time. (B) Rates of Herceptin-dependent killing of the SKBR3 cells by CD16.NK92 effectors in the absence (black bar) or presence of either TS2/4 (white bar) or R6.5 (shaded bar) antibodies are shown. The killing rates were calculated from initial slopes for the killing curves presented in panel A. The data represent pooled average of 2 independent experiments and are shown as mean±sd. P values were calculated by Student s t test: ***p<0.001; ns, not significant.
Figure S2. Level of expression of ICAM-1 influences killing kinetics of SKBR-3 target cells. Specific lysis of SKBR-3 cells transfected with ICAM-1 sirna (diamonds) or control sirna (filled circles) is shown. Cr 51 -labeled SKBR3 cells were combined with CD16.NK-92 cells at 1:5 ratio and incubated with 1 µg/ml of Herceptin antibody for indicated periods of time. The data are representative of two independent experiments and are shown as mean±sd. P values were calculated by Student s t test: *p<0.05, **p<0.01, ***p<0.001; ns, not significant. SKBR3 cells were grown to 80% of confluence in a 6-well plate. On the day of sirna transfection, 150 pmol of control sirna or ICAM-1 sirma (Santa Cruz Biotecnology) were diluted with 125 µl of Opti-MEM medium and mixed with equal volume of the medium containing 10 µl of Lipophectamine 3000 (Life Technology). After 5 min of incubation, the mixtures were added to the wells containing cells. The plate were incubated for 6 hours at 37 o C in CO 2 incubator, and the transfection medium was replaced with 2 ml of complete SKBR3 culture medium containing 100 u/ml of IFN-γ. After overnight incubation, the cells were used as a target in the cytolytic assay as described in Fig. S1 or were analyzed by Flow Cytometry to test the expression of ICAM-1. 2 10 5 SKBR3 cells were sequentially incubated with anti-icam-1 antibody R6.5 (5 µg/ml) and a secondary anti-mouse antibody labeled with Alexa Fluor 647 (0.1 µg/ml) for 30 minutes at 4 C. The cells were then washed free of unreacted reagents and analyzed by LSR II Flow Cytometer. Treatment of SKBR3 cells with ICAM-1 sirna resulted in 1.3 times decrease of the surface level of ICAM-1 (data not shown).
Figure S3. Effect of differential expression of ICAM-1 on SKBR-3 cells on their ability to form conjugates with CD16.NK-92 effectors. Intact (filled bars) or IFN-γ treated (dashed bars) SKBR3 cells were stained with calcein and incubated with PKH26-labeled CD16.NK92 cells in the presence of 10µg/ml Herceptin for indicated periods of time at 37 o C. The samples were then fixed with paraformaldehyde and CD16.NK92/SKBR3 conjugates were identified and counted as double-stained pairs by Flow Cytometry. Results are representative of 3 independent experiments.
Figure S4. Blocking of integrin-mediated signaling with Pyk2 inhibitor impairs killing of SKBR3 target cells in a dose-dependent manner. Specific cell lysis of SKBR-3 cells by either intact (black bar) or treated with Pyk2 inhibitor (all other bars) CD16.NK-92 is shown. Cr 51 -labeled SKBR3 cells were combined with treated or untreated CD16.NK-92 cells at 1:5 ratio and incubated with 1 µg/ml of Herceptin antibody. (A) CD16.NK-92 cells were treated with Pyk2 inhibitor at 10 µm for 30 minutes. The inhibitor was either kept in extracellular medium during the cytolytic assay (dashed bar) or washed away prior to the assay (white bar). The percent specific lysis was determined after 4 hours incubation at 37 0 C and 5% CO 2. The treated CD16.NK-92 cells completely restored their activity after an excess of the inhibitor was removed (white bar). Thus, the observed inhibitory effect was due to the inhibition of integrin-mediated signaling and was not caused by other Pyk2-mediated effects including death of CD16.NK-92 cells. The latter was also evident from Flow Cytometry analysis the cells stained with propidium iodide (data not shown). (B) CD16.NK-92 cells were either treated or untreated with Pyk2 inhibitor at indicated concentrations for 30 minutes, and the cells were used in the cytolytic assay as described above. Percent specific lysis was measured after one hour of incubation to match the time for the analysis of inhibition of the size and dynamics of microclusters (Fig.4 and 5). The observed inhibitory effect was dose dependent and more profound for the killing of IFN-γ treated target cells. The latter is due to upregulation of ICAM-1 accounting for a larger fraction of engaged integrin receptors on the effector cells that enhances the cytolytic activity and increases the inhibitory effect. Data are representative of two independent experiments and are shown as mean±sd. P values were calculated by Student s t test: *p<0.05; **p<0.005; ***p<0.0001; ns, not significant.
Figure S5. Kymographs showing continuous degranulation of CD16.NK-92 cells. The cells were exposed to planar lipid bilayers that display anti-cd16 antibodies in the presence or absence of ICAM-1. Granule release at the cell/bilayer interface was observed for 1.5 hours after the initial cell contact with the bilayers using TIRF and wide field fluorescence microscopy. Sequential images of CD16.NK-92 cell/bilayer interface showing appearance of CD107 as a marker of the cytolytic granule release were collected and composed into stacks using MetaMorph image processing software s 3D view function to produce an image of temporalspatial distribution of cytotoxic granules. (A) Real time degranulation of CD16.NK-92 cells on planar bilayers that do (upper panel) or do not (lower panel) display ICAM-1 molecules. Representative kymographs show temporal changes in spatial position of cytolytic granules released at the cell/bilayer interface during 90 minutes. Granule release locations are shown in red. X and Y coordinates indicate the position of the cell/bilayer interface (blue), and time axis indicates temporal changes in the spatial position of cytolytic granule locations; earlier time points are on the left and later time points on the right. (B) Relative positioning of degranulation foci (red) and ICAM-1 molecules (blue) at the contact area of CD16.NK-92 at different time points. Representative images show cytotoxic granule release by a CD16.NK-92 interacting with bilayers containing anti-cd16 antibodies in the presence of ICAM-1. ICAM-1 accumulation locations (blue) were visualized by wide field fluorescent microscopy; granule release locations (red) were imaged by TIRF microscopy.
Figure S6. Effect of integrin engagement on molecular events at the CD16.NK-92/bilayer interface. CD16.NK-92 cells were exposed to planar lipid bilayers containing anti-cd16 antibodies with (upper bars) or without (lower bars) ICAM-1. Cell/bilayer interactions were observed for 1.5 hours after initial cell-bilayer contact using TIRF and wide field fluorescent microscopy. Colored bars define average periods of time during which accumulation of different molecules was observed at CD16.NK-92/bilayer interface. Ligands incorporated into the bilayers are shown on the left. Green bars: CD16 microcluster formation. Blue bar: ICAM-1 accumulation. Red bars: cytolytic granule release measured by the appearance of CD107a at the interface. Results are representative of 4 independent experiments.
Figure S7. Formation and centripetal movements of CD16 microclusters at the contact surface of CD16.NK-92 cells. The cells were exposed to planar lipid bilayers containing anti- CD16 antibodies with (upper row) or without (lower row) ICAM-1. After the cells were allowed to adhere to the bilayers, TIRFM images were taken every 30 seconds for the entire time of observation. The interface images at indicated time points are shown. Movie 1 and 2 show the dynamics of the events.
Figure S8. Identification of single microclusters. To identify single microclusters participating in centripetal movement prior to their merger with other microclusters, lines were drawn in several directions across the microclusters (left panel) and the fluorescent intensity profiles along these lines were examined. Microclusters demonstrating unimodal intensity profiles (lower right panel), such as microclusters 3 and 6 on the left panel, were identified as single microclusters and included in the analysis. Microclusters demonstrating bi- or polymodal intensity profiles (upper right panel), such as 1 and 5, were identified as merged microclusters and were excluded from the analysis.
Figure S9. Size of CD16 microclusters was preserved over time and was dependent on the presence of integrin receptor ligand ICAM-1 rather than on CD16 receptor ligand density. CD16.NK-92 cells were exposed to bilayers containing anti-cd16 antibody at 5 or 50 or 500 mol/µm 2 with or without ICAM-1. CD16 microclusters formation was analyzed after 5, 10 or 30 minutes followed by initial cell/bilayer contact. Integrated fluorescent intensities of individual microclusters were determined. (A) The distribution of the integrated fluorescent intensities at the contact area of CD16.NK-92 cells exposed to bilayers containing anti-cd16 antibody at 500 molecules/µm 2 (blue) or 50 molecules/µm 2 (red) in the presence of ICAM-1 (300 molecules/µm 2 ) (B) The distributions of integrated fluorescent intensities of individual microclusters at the contact area of CD16.NK-92 cells exposed to bilayers containing anti-cd16 antibody (50 molecules/µm 2 ) in the presence (blue) or absence (red) of ICAM-1. (C) Average integrated fluorescent intensities of CD16 individual microclusters at CD16.NK-92/bilayer interface in the presence (filled bars) or absence (empty bars) of ICAM-1 determined at indicated time points after the initial cell/bilayer contact. (D) Average integrated fluorescent intensities of individual microclusters observed at the contact area of CD16.NK-92 cells on bilayers containing anti-cd16 antibodies at indicated density in the presence (filled bars) or absence (empty bars) of ICAM-1. At least 100 microclusters in each group were analyzed. Results are representative from 4 independent experiments. *p<0.05 by Student s t-test, ns no significant statistical difference by Student s t-test.
Figure S10. The parameters of single CD16 microcluster movement. To measure the indicated parameters of microcluster movement, images of CD16 microclusters were taken every 10 seconds. The change in location for a microcluster between two frames was assumed to be significant if the distance between coordinates of the microcluster s center in those frames exceeds 0.32 µm (2 pixels). The beginning of the movement was defined as the last frame where locations of all microclusters were not changed significantly from the previous frame. The end of movement was defined as the last frame in which changes in locations for all microcluster centers from the previous frame were still significant. At the beginning of its movement, at time point t 0, each microcluster was located at point p 0i with coordinates x 0i and y 0i. At the end of movement, t m, microclusters were located in point p mi with coordinates x mi and y mi. Microcluster displacement, d, for the i th microcluster was determined as the shortest distance between the initial and the final position of the microcluster: d i = p mi p 0i Microcluster mobility was determined as the period of time between the beginning of microcluster movement and the end of the movement of all microclusters at the contact surface of a cell: τ = t m t 0
Movie Legends Movie 1. Formation and centripetal movements of CD16 microclusters at interface between a CD16.NK-92 cell and a planar lipid bilayer containing anti-cd16 antibodies and ICAM-1. After the cell was allowed to adhere to the bilayer, TIRFM images were taken every 30 seconds for the entire time of observation and then montaged into the movie using Metamorph software. Movie 2. The formation and centripetal movements of CD16 microclusters at an interface between a CD16.NK92 cell and a planar lipid bilayer containing anti-cd16 antibodies in the absence of ICAM-1. The movie was montaged as for Movie 1. Movie 3. Relative positioning of CD16 microclusters and cytotoxic granules at interface between a CD16.NK92 cell and a planar lipid bilayer containing anti-cd16 antibodies and ICAM-1. The movie was assembled as Movie 1. Microclusters are in green, cytotoxic granules are in red. Movie 4. Relative positioning of CD16 microclusters and cytotoxic granules at an interface between CD16.NK92 cell and planar lipid bilayer containing anti-cd16 antibodies in the absence of ICAM-1. The movie was assembled as for Movie 1. Microclusters are in green, cytotoxic granules are shown in red. Movie 5. Tracking of single CD16 microcluster movements at an interface between a CD16.NK92 cell and a planar lipid bilayer containing anti-cd16 antibodies and ICAM-1. After the cell was allowed to adhere to the bilayer, TIRFM images were taken every 10 seconds for the entire time of observation. Using the track object function of the MetaMorph software, several single microclusters were tracked from the moment they started moving till the moment they merged with another microcluster. Individual frames were subsequently montaged into the movie.