Abstract

Abstract 3537

Poster Board III-474

Natural killer (NK) cells have been shown to attack virally-infected and transformed cells as well allogeneic bone marrow cells (BMC) but not normal self-tissues. The mechanism of missing self recognition and self tolerance of NK cells is poorly understood. NK cells exist as subsets based on expression of inhibitory receptors (Ly49 in mouse, KIR in man) that bind MHC class molecules. In vitro data have shown that murine NK cell subsets bearing Ly49 receptors for self MHC class I molecules have intrinsically higher effector function, supporting the hypothesis that NK cells undergo a host MHC class I-dependent functional education, allowing the NK cells bearing the appropriate Ly49 receptors to preferentially mediate effector function. Thus far, no in vivo evidence for this preferential licensing or arming has been shown. We assessed the intrinsic response potential of the different Ly49+ NK cell subsets in BMC rejection without having the complicating effects of binding MHC on the target cell (which delivers potent inhibitory signals to the NK cell) by using β2-microglobulin deficient (β2m−/−) mice as donors which totally lack class I MHC molecules. Using syngeneic, congenic, and allogeneic strains of mice as recipients and depleting the different Ly49 subsets, we found that NK cell subsets whose Ly49 molecules have been shown to bind “self-MHC Class I” (ie Ly49G2 in H2d and Ly49C in H2b haplotypes) were found to be the dominant subset responsible for mediating the rejection of the β2m−/− BMC. This provides the first in vivo evidence for host MHC class I-dependent functional education (licensing or arming). Importantly, we also demonstrated that prior activation of the NK cells in vivo could override this licensing effect and allow the “non-licensed subset” to mediate rejection. The pattern of NK mediated rejection ability by Ly49 subsets was observed in B10D2 allogeneic H2d strain mice but not observed in BALB/c allogeneic H2d strain mice indicating that licensing ability was not solely dependent on host MHC expression and supporting the role of MHC class I–independent system for NK-cell self-tolerance. Interestingly, all H2d strain mice (B10.D2, BALB/c, and B6D2F1) were able to resist significantly greater amounts of B2m-/- BMC than H2b strain mice indicating that the rheostat hypothesis regarding Ly49 affinities for MHC and NK cell function impacts BMC rejection capability. These results demonstrate that both MHC and non-MHC genes on the host has multiple effects on NK cell subset-mediated BMC rejection and that licensing or arming of Ly49 NK cell subsets can be observed in vivo.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.