Engineering the triple mutation, DLE, into the CH2 domain of the anti-CD33 antibody improves NK cell-mediated ADCC. (A-B) Sandwich ELISA confirmed that the affinity of Fc-engineered DLE-HuM195 for CD16 is increased, relative to wt-HuM195, with no change in antigen affinity (CD33). (C-E) Population-level calcein AM release assays profiling cytotoxicity of ex vivo NK cells against mAb-precoated AML cell lines at varying E:T ratios and a fixed concentration of mAb (1 µg/ml). N = 3 donors, run in triplicate, for 6 hours. (F-H) Population-level calcein AM release assay run against CD33⁺EL4 target cells incubated with mAb. NK cell-mediated lysis when targets were precoated with mAb (excess mAb is washed away after preincubation) at (F) a fixed E:T ratio, but varying mAb concentration, and (G) fixed mAb concentration but varying E:T ratios. (H) NK cell-mediated target lysis when mAb is present in solution for the duration of the assay. Frequencies are reported with background correction (values obtained without mAb). (I) Single-cell endpoint cytotoxicity assay. Labeled CD33⁺EL4 target cells were precoated with mAb (1 µg/mL) and incubated on a nanowell array with labeled, expanded NK cells as effectors for 6 hours, and ADCC was determined as described in supplemental Figure 3 (E:T ratio, 1:1-3) (at least 2676 events were analyzed for each donor and mAb condition was tested). Of note, these values represent the background-corrected frequencies of cytolytic NK cells, which are obtained as described in supplemental Figure 3. When available, CD16-158 allotype is annotated near the donor identifying number (F, phenylalanine; V, valine). All data are shown as mean ± SD, and stars represent P values calculated as detailed in supplemental Table 1 (P₁-P₇). nd, not determined.