NK cells of XLP2 patients had impaired natural cytotoxicity and developed adaptive-like phenotype. (A) Cytotoxic activity of NK cells in patients 1 and 2 at 24 years of age, when they were diagnosed with XLP2 (see timeline in Figure 1). Shown is the result of the ⁵¹Cr-release assay (for clarity, error bars are not shown). (B) Memory B-cell phenotype (CD3⁻CD19⁺C20⁺CD27⁺) at the first onset of HLH, 6 years later (at the time of XLP2 diagnosis, aged 24 years), and 1.5 years after the diagnosis (aged 25.5 years). (C) Cytotoxic activity of the NK cells of patients 1 and 2, 1 year after XLP2 diagnosis. Shown are the ⁵¹Cr-release assay (for clarity, error bars are not shown) and the CD107a degranulation assay. (D) Cytotoxic activity of CD8⁺ T cells (please see supplemental Methods for details) as measured by ⁵¹Cr-release assay (mean ± standard error of the mean [SEM], n = 3) and CD107a degranulation assay. Tests were done at the same time as in panel C. (E) NK cells of patient 1 and 2, but not patient 3, developed adaptive-like phenotype (for gating strategy, please see supplemental Figure 4). NK cells in patients 1/2 were assessed 1 year (as in panel C) and 1.5 years after the diagnosis (depicted by gray and blue bars, respectively), and in patient 3 – 1.75 years after the diagnosis. Shown is mean ± standard deviation (SD) of 10 (FcεRγ⁻) or 11 (NKG2C⁺ and NKG2A⁺) controls. There was a similar trend for NKG2A⁺ and NKG2C⁺ cells, when gated on both CD16⁺CD56^high and CD16⁺CD56^dim NK cells (data not shown). (F) NK cells from patients 1 and 2 show enhanced response to rituximab (RTX)-treated Raji cells through ADCC. E/T, effector/target ratio.