Genetic and Mechanistic Dissection of Missing-Self Killing by Natural Killer Cells

Natural killer (NK) cells eliminate transformed and infected cells through activating receptor-mediated cytotoxicity and through recognition of cells that have lost major histocompatibility complex class I (MHC-I) expression, a process termed missing-self recognition. Although missing-self recognition is a central principle of NK-cell biology, the mechanisms that execute missing-self killing remain incompletely understood. In this dissertation, we used complementary genetic and functional approaches to investigate the molecular basis of missing-self killing. We first implemented a CRISPR–Cas9 ribonucleoprotein electroporation platform for efficient genetic manipulation of primary murine NK cells, providing a versatile approach for functional studies of NK-cell biology. To identify determinants of target susceptibility to missing-self killing, we next used CRISPR-engineered TAP2-deficient RMA cells together with a genome-wide CRISPR resistance screen. The screen identified only two reproducibly enriched genes: TAP2 and intercellular adhesion molecule-1 (ICAM-1). Restoration of TAP2 expression re-established MHC-I surface expression and conferred resistance to NK-cell killing, validating the specificity of the screening strategy. Loss of ICAM-1 similarly conferred resistance to missing-self killing and was confirmed by genetic deletion and antibody blockade. In contrast, ICAM-1 was dispensable for Ly49H-mediated killing of m157-expressing targets. Although CD11a deficiency modestly impaired missing-self killing in vitro, neither CD11a deficiency nor loss of target-cell ICAM-1 impaired missing-self rejection in vivo, suggesting the existence of compensatory mechanisms. Consistent with this possibility, mice lacking individual NK-cell signaling adaptors, including DAP10, DAP12, FcRγ, or CD3ζ, retained normal missing-self rejection, whereas simultaneous loss of DAP10, DAP12, and FcRγ abolished killing of ICAM-1-deficient missing-self targets. Mechanistic studies further revealed a dissociation between killing and conventional degranulation: missing-self targets failed to induce detectable bulk NK-cell degranulation, yet killing remained largely dependent on perforin, suggesting that missing-self recognition engages a distinct mode of cytotoxicity not captured by conventional activation assays. Together, these findings identify ICAM-1 as a dominant non-MHC determinant of susceptibility to missing-self killing and support a model in which missing-self cytotoxicity is maintained through redundant contributions from ICAM-1-mediated recognition and adaptor-dependent signaling pathways.