GzmB binding to heparan sulfate is dispensable for NK cell effector function. Apoptosis induction of HL-60 and K562 cells by the human NK92 cell line was assessed by ³H-thymidine release. (A) To block binding sites on target cells, HL-60 (top) and K562 cells (bottom) were preincubated with GzmB^S195A, which was then kept constant at 25 μg/mL during the entire killing experiment (3.5 hours). indicates medium; ×, GzmB^S195A; □, EGTA/MgCl₂. (A-B) EGTA/MgCl₂ (6 mM/3 mM) and the pan-caspase inhibitor z-Val-Ala-Asp-Fluoromethylketone (z-VAD-fmk) (50 μM) were added in control experiments to verify perforin and caspase involvement. (B) Apoptosis induced by NK cell–delivered GzmB is independent of GzmB cell surface binding. To abrogate GzmB binding, we incubated HL-60 cells for 36 hours in 50 mM sodium chlorate. Thereafter, sodium chlorate concentrations were maintained at 10 mM (3.5 hours) (indicated by ×). As a control, nontreated HL-60 cells were coincubated with NK92 cells in the presence of 10 mM sodium chlorate (○). indicates medium; □, EGTA/MgCl₂; and ▵, z-VAD-fmk. (C) Disappearance of GzmB-binding sites on sodium chlorate treatment. HL-60 cells were treated with sodium chlorate at 50 mM for 36 hours and then at 10 mM for 3.5 hours (□) and were compared with cells treated only for 3.5 hours at 10 mM (▦). ▪ indicates medium. Binding of GzmB-Bio and the 10E4 HS epitope was detected with streptavidin-PE (SA-PE) and FITC-labeled goat antimouse antibodies (GaM-FITC) by FACS analysis.