Natural killer cell tolerance : influence of the MHC class I expression in the host

Development of self-tolerance in the immune system is characterized by somatic processes aiming at permanently eliminating, or silencing, potentially autoreactive lymphocytes. Such tolerance mechanisms are of outmost importance if autoimmunity is to be avoided. This thesis deals with tolerance development in natural killer (NK) cells, the third largest lymphocyte population. These cells are part of the innate immune system and participate in rejection of tumor cells and bone marrow transplants, clearance of viral and other infections, and in immunoregulation. NK cells were discovered in 1975, and were at that time considered to be a homogenous population of cells capable only of nonspecific effector mechanisms. As the role of the polymorphic MHC class I molecules in NK cell recognition has gradually emerged, there has been an increasing interest in exploring somatic processes for shaping of NK cell specificity and tolerance.

The NK cell pool is now known to contain a large number of subpopulations with different specificities. The specificity of each NK cell is determined by its composition of MHC-specific receptors, both inhibitory and activating. In the mouse, those are either Ly49 or CD941NKG2 receptors. Each Ly49 receptor is specific for a set of MHC class la molecules, while CD94/NKG2A, the only receptor in this family whose specificity is known, recognizes an MHC class lb molecule, Qa-1. NK cell specificity is a complex issue, since most NK cells express more than one, perhaps an average of three to four, different receptors, both activating and inhibitory ones. The studies included here describes attempts to dissect the complexity of NK cell specificity and tolerance, in particular how host MHC class I molecules influence NK cells. MHC class I transgenic mice have been central tools, including one line with a mosaic expression pattern of the introduced gene. Two different cellular phenotypes, sharing two MHC class I alleles but differing in expression of a third, coexist in these mice.

The studies have revealed that the peptide-binding part of host MHC class I molecules strongly influence development of NK cell specificity. However, not all MHC class I molecules of the host have a critical influence on NK cell education. With regard to self-tolerance, the studies with mice containing mixed MHC phenotypes have shown that autoreactive NK cells are not eliminated, at least not all of them: when tolerant NK cells were removed from their mosaic environment and cultured in vitro, tolerance was rapidly lost. This suggests that NK cell tolerance, in contrast to central T cell tolerance, may rely on mechanisms that need to operate continuously in vivo. Studies in mixed bone marrow chimeric mice indicated that the number of interactions between NK cells and potential targets with a given MHC phenotype may be important. The studies on Ly49A receptors in mosaic mice showed that the MHC class I molecules on surrounding cells as well as on the NK cell itself play distinct roles in regulating receptor expression. Further studies are required to determine how this relates to tolerance. Finally, liver and intestinal expression of an MHC class I transgene was sufficient to induce T-cell tolerance, but did not influence NK cell specificity.

In conclusion, although NK cell specificity is determined by the peptide presenting part of the MHC class I molecule, the processes leading to tolerance may be fundamentally different from those regulating the T-cell repertoire. Continuous regulation through interactions with surrounding cells is discussed as one possibly important factor in NK cell tolerance.