Hybrid resistance and the Ly-49 family of natural killer cell receptors

T he phenomenon of hybrid resistance occurs when irradiated F1 hybrid mice reject bone marrow cells (BMC) donated by either parent (1). Hybrid resistance is of major interest in immunology because it violates the classical laws for solid tissue transplantation, upon which the MHC was defined. (F1 hybrid mice normally accept solid tissue donated by either parent.) Recent studies, including those published in this issue (2, 3), have yielded significant insight into the relationship between NK cells and this perplexing biological phenomenon, providing a means to determine the molecular basis for NK cell specificity for their targets. Initially described on the basis of an apparently innate ability to kill certain tumor cells, termed natural killing, NK cells are now also known to be surface immunoglobulin-negative lymphocytes that are important in early host defense against infections and in regulating the subsequent specific immune response to intracellular pathogens by producing cytokines (4). NK cells closely resemble T cells with regard to display of cell surface molecules, cytolytic mechanisms to kill their cellular targets, and cytokine production profiles. It is not surprising, therefore, that they may share an immediate common progenitor (5, 6). NK cells clearly must use a mechanism to discriminate between cells that they kill and those that they spare, such as normal cells. However, NK cells differ from T cells in target recognition mechanisms in two important ways. First, NK cells do not express the TCR/CD3 complex (7). Indeed, NK cells do not rearrange TCR genes and appear to be normal in mice with genetic abnormalities in TCR gene recombination, such as scid, and targeted RAG-1 or RAG-2 deficiency (8-10). Second, by contrast to MHC class I-restricted T cells, NK cells do not require the expression of MHC class I molecules on targets for natural killing (11). In fact, NK cells kill better when their targets do not express MHC class I. To explain this, K~irre postulated the "missing self" hypothesis, suggesting that NK cells survey tissues for MHC class I expression that is nearly ubiquitously expressed (12). With abnormal expression of MHC class I, the NK cell is released from this inhibitory influence and the target is killed. Certainly, there is now ample in vitro and in vivo evidence to support the general tenet of the hypothesis that MHC class I inhibits target killing. Although these findings were initially controversial because they were difficult to reproduce, more recent studies have shown that not all MHC class I molecules were capable of conferring resistance to targets. Selective resistance has been surprisingly mapped to the peptide binding domains of MHC class I molecules (13), suggesting that NK cells may discriminate between MHC class I molecules by recognition of the MHC class I epitopes that are also recognized by T cells, but in a TCR-independent manner. In hybrid resistance, NK cells have been identified as the host elements that reject parental BMC (1). Hybrid resistance can occur in the apparent absence of immunoglobulin and T cells because it occurs in hosts with the scid mutation. Although the presumption is that NK cells are directly cytotoxic against BM precursors, the ability of NK cells to produce cytokines, such as interferon-3,, TNF-cr GM-CSF, etc. (4), which may influence bone marrow engraftment raises the possibility that differential cytokine production may be another mechanism. Nevertheless, even this latter possibility suggests that NK cells must be specifically triggered (or inhibited, see below) to manifest hybrid resistance. Parental Determinants Recognized by Host N K Cells. Recipient NK cells mediate BMC rejection by recognizing parental determinants encoded by genes that have been mapped to the MHC (1). Moreover, an H-2 b mouse transgenic for H-2D d resembles the (H-2 b x H-2d)F1 hybrid with regard to hybrid resistance (14). In addition, host NK cells in otherwise syngeneic hosts reject MHC class I-deficient BMC derived from mice with a targeted mutation in the ~2microglobulin gene (15, 16). This rejection is thus highly reminiscent of hybrid resistance. Taken together, the data strongly suggest that NK cells in F1 hybrid animals mediate hybrid resistance against parental BMC cells that do not express the full complement of MHC class I molecules that would be codominantly expressed on F1 hybrid cells. Other considerations, however, have prevented a consensus on this interpretation. The Bennett and Kumar group suggest that NK cells recognize parental determinants termed "hematopoietic histocompatibility" (Hh-1) antigens that can be distinguished from MHC molecules for several reasons (1). Their extensive studies of MHC-congenic mice demonstrated that the specificity of rejection could not be conveniently grouped according to MHC class I haplotypes of either the donor or recipient (17). Moreover, there is significant overlap between the groups. This analysis, however, has required interpretation without knowledge of putative NK cell receptors that presumably also affect the specificity. Analyses of MHC recombinant congenic mice have also suggested that