The detection of aberrant cells by natural killer (NK) cells is controlled by the integration of signals from activating and inhibitory ligands and from cytokines such as IL-15. We identified cytokine-inducible SH2-containing protein (CIS, encoded by Cish) as a critical negative regulator of IL-15 signaling in NK cells. Cish was rapidly induced in response to IL-15, and deletion of Cish rendered NK cells hypersensitive to IL-15, as evidenced by enhanced proliferation, survival, IFN-γ production and cytotoxicity toward tumors. This was associated with increased JAK-STAT signaling in NK cells in which Cish was deleted. Correspondingly, CIS interacted with the tyrosine kinase JAK1, inhibiting its enzymatic activity and targeting JAK for proteasomal degradation. Cish(-/-) mice were resistant to melanoma, prostate and breast cancer metastasis in vivo, and this was intrinsic to NK cell activity. Our data uncover a potent intracellular checkpoint in NK cell-mediated tumor immunity and suggest possibilities for new cancer immunotherapies directed at blocking CIS function.

The cytokine IL-15 is required for natural killer (NK) cell homeostasis; however, the intrinsic mechanism governing this requirement remains unexplored. Here we identify the absolute requirement for myeloid cell leukaemia sequence-1 (Mcl1) in the sustained survival of NK cells in vivo. Mcl1 is highly expressed in NK cells and regulated by IL-15 in a dose-dependent manner via STAT5 phosphorylation and subsequent binding to the 3 0 -UTR of Mcl1. Specific deletion of Mcl1 in NK cells results in the absolute loss of NK cells from all tissues owing to a failure to antagonize pro-apoptotic proteins in the outer mitochondrial membrane. This NK lymphopenia results in mice succumbing to multiorgan melanoma metastases, being permissive to allogeneic transplantation and being resistant to toxic shock following polymicrobial sepsis challenge. These results clearly demonstrate a non-redundant pathway linking IL-15 to Mcl1 in the maintenance of NK cells and innate immune responses in vivo.

Survival of various immune cell populations has been proposed to preferentially rely on a particular anti-apoptotic BCL-2 family member, for example, naive T cells require BCL-2, while regulatory T cells require MCL-1. Here we examined the survival requirements of multiple immune cell subsets in vitro and in vivo, using both genetic and pharmacological approaches. Our findings support a model in which survival is determined by quantitative participation of multiple anti-apoptotic proteins rather than by a single anti-apoptotic protein. This model provides both an insight into how the sum of relative levels of anti-apoptotic proteins BCL-2, MCL-1 and A1 influence survival of T cells, B cells and dendritic cells, and a framework for ascertaining how these different immune cells can be optimally targeted in treatment of immunopathology, transplantation rejection or hematological cancers.