NK cells are necessary and sufficient to control EAU through production of IFN-γ. (A) CD45.2 WT recipients were or were not treated with anti–NK1.1. After 2 d, mice were challenged for EAU and repleted with CD45.1 WT NK cells. The gating strategy is shown in A and the numbers of CD45.2 (host) and CD45.1 (donor) NK cells were determined from the draining LNs 5 d after immunization (A). (B–D) GKO mice were depleted of endogenous NK cells by anti-NK1.1 antibody. 2 d later, mice were repleted with WT or GKO NK cells and immunized for EAU. (B) Disease was monitored at the indicated times. (C and D) Expression of IL-17A and GM-CSF in eye-infiltrating CD4⁺ cells of mice in B, as determined by intracellular staining on day 21. (E) GKO mice were immunized for EAU and WT or GKO donor NK cells were infused on the same day. Disease was monitored at the indicated times. IFN-γ expression in the draining LNs was determined in CFSE-labeled donor cells by intracellular cytokine staining on d 5. (F) Mice were given IFN-γ–neutralizing antibody or isotype control. (G) Mice received WT or GKO NK cells on day 0 or day 9 after immunization. (H) IL-15 KO mice received WT or GKO NK cells on day 0 after immunization. Data are combined from 2 (E–H) or 3 (B and C) independent experiments totaling at least 12 mice per point, or are representative of 2 independent experiments with at least 3 mice per group (A and D). *, P < 0.05; (B and F–H) linear regression or (D) Mann-Whitney U test.