NK cell recognition : adaptability to host factors in normal and diabetic mice

Abstract

The overall aim of this thesis was to understand the rules, which guide NK cell behavior in vivo. NK cells are part of the innate immune system and are important in defense against certain virus infections and tumorigenic cells. They also play a role in rejecting bone marrow grafts.

In the first study, we further characterized an earlier reported activation defect in NK cells from Non-Obese Diabetic (NOD) mice. This defect was found to have a broad specificity, and affect several proximal activation pathways. The defect was however partially restorable with the proper stimulation of the NK cells; a combination of IL-12 and IL-18 in vitro or the interferon inducer tilorone in vivo. This suggested that NOD NK cells have the capacity to perform cytotoxic activity if stimulated properly.

In the second study, we further developed a technique for measurement of in vivo cytotoxicity based on labeling of target cells with the intracellular dye CFSE. We established protocols for measuring killing of normal lymphocytes as well as tumor targets. With this method, we could for the first time verify that resting, non-proliferating lymphocytes were killed according to the missing self principle. The missing self hypothesis states that NK cells are able to react on the lack of expression of self MHC class I proteins on target cells by killing the target cells. This missing self reactivity was independent of DAP12 signalling. Killing of allogeneic targets was DAP12 dependent in naive targets, but DAP12 independent in recipient mice activated with tilorone. This suggested that the relevant activation pathways changed with the activation stimulus.

In the third study, we explored the missing self principle more extensively by examining the role of different MHC class I proteins for missing self reactivity. Individual MHC class I proteins was shown to evoke missing self responses of different strengths, which was defined as the educating impact of the MHC class I protein. The educating impact of weak MHC class I ligands was further attenuated by the introduction of additional MHC class I proteins, whereas MHC class I proteins with a strong educating impact maintained their efficiency under such conditions. This showed that the impact of a certain MHC class I protein can differ according to other MHC class I proteins expressed in the host.

In the fourth study, we studied the effect on diabetes development in NOD mice by treating the mice with a depleting antibody against IL-2Rbeta. IL-2Rbeta is expressed at high levels on all NK cells, and a small subset of CD8+ T cells. This treatment completely protected against diabetes development in a long-term treatment protocol and was mediated by intervening with a late event in the disease progress. A protection that is effective late in the disease course is interesting with regard to a treatment in human patients. Since the anti-IL-2Rbeta treatment depleted both NK cells and a subset of CD8+ T cells, the characterization of the IL-2Rbeta positive cell type responsible for the disease promotion has to await further studies.

In conclusion, this thesis points towards a dynamic control of NK cell behavior in vivo, which is influenced by the internal cytokine milieu in the host. Activation of NK cell killing can proceed via several pathways depending on the preceding activation stimulus. Results discussed in this thesis also suggest that NK cells can calibrate their activation status according to the inhibitory signals they receive from the MHC class I proteins expressed by the host.