Our studies appear to provide a unifying explanation for how natural killer (NK) cells achieve tolerance to self-tissues, a central question in NK cell biology. NK cells have been enigmatic because these innate immune cells appear to distinguish between diseased and normal cells. Diseased cells, such as infected or tumor cells, are often killed by NK cells and activate NK cells to produce inflammatory cytokines. Such responses could be detrimental to normal cells unless NK cell activation is controlled.

“Our studies appear to provide a unifying explanation for how natural killer (NK) cells achieve tolerance to self-tissues, a central question in NK cell biology.” Wayne M. Yokoyama, photo ©Paul Fetters.

Prior studies on this broad topic of NK cell "tolerance" revealed that NK cells express a plethora of germline-encoded, non-rearranged receptors that recognize ligands on their targets and either activate or inhibit NK cell responses. One major set of receptors recognizes major histocompatibility complex (MHC) class I molecules on targets and inhibit NK cell activation. These receptors explain Professor Klas Kärre’s (of the Karolinska Institute) "missing-self" hypothesis, suggesting that NK cells survey tissues for normal, ubiquitously expressed MHC class I molecules.

In the absence of MHC class I, which occurs during tumorigenesis or viral infection, NK cells can then kill the target and produce cytokines. However, prior studies also showed that NK cells do not attack otherwise normal tissues in hosts lacking MHC class I. Indeed, NK cells are generally "hyporeactive" when isolated from MHC-deficient hosts and tested against tumor target cells.

Moreover, the genes for the NK cell receptors and MHC class I are highly polymorphic yet are encoded on different chromosomes, indicating the presence of a mechanism to endow the NK cell with the appropriate tolerance to self. Our studies showed that NK cells become functionally competent through an interaction with their MHC-specific receptor and self-MHC.

We used a deceptively simple assay to directly determine the functional capacities of individual NK cells ex vivo. Most prior studies of NK cells focused on the ability of bulk populations of activated NK cells to kill tumor targets in a ⁵¹Cr-release assay. However, the universe of receptors involved in killing of tumors is still incompletely understood.

Here we triggered freshly isolated NK cells with an immobilized antibody that recognizes an activation receptor (NK1.1 or Nkrp1c) on all immature and mature NK cells. Furthermore, we determined the response of individual NK cells by flow cytometry for intracellular interferon-y (IFNy) production.

This assay led to the striking observation that there is a defect in activation receptor triggering in individual NK cells from MHC-deficient mice, indicating that MHC class I normally "licenses" NK cells to become functionally competent. Furthermore, we showed that the ability to be triggered was directly related to the presence of an NK cell receptor (belonging to the Ly49 family) for self-MHC class I.

With a mouse transgenic for a single MHC class I molecule, and gene transfer of the NK cell receptor, we further demonstrated that licensing of NK cells was due to signaling by the NK cell receptor itself, indicating a second function of these NK cell receptors. Ironically, the Ly49 receptors that confer the licensing process were originally identified because they inhibit NK cell activation in effector responses.

Our studies indicate that NK cells become functionally competent after their receptors for MHC class I are engaged by self-MHC. These cells are tolerant for self because they have receptors that bind self-MHC and inhibit NK cell effector responses; these receptors are the same ones that license NK cells.

There is another type of self-tolerant NK cells, those that do not express receptors for self-MHC. They are functionally incompetent. Thus, our studies provide insight into how NK cells acquire the ability to recognize and spare normal self-tissues.

Although we remain interested in the molecular bases for NK cell functions and differentiation, our recent work on licensing has led us to studies of human NK cells. Our preliminary data strongly suggest that MHC class I-dependent licensing also applies to human NK cells. We feel this is an important advance since others have noted a correlation between outcomes of certain diseases and an individual’s complement of NK cell receptor and HLA genotypes. Licensing provides another potential explanation for these epidemiological data.

The possible contribution of licensed NK cells to disease outcomes could lead to new insights for therapeutic interventions.