Functional NK cell deficiencies are associated with autoimmune diseases, including multiple sclerosis (MS). NK cells can promote or inhibit adaptive immunity via either cytokine production or cytotoxicity towards immature dendritic cells and activated T cells. In humans, this immunoregulatory role resides in the CD56bright NK cell subset, which is selectively expanded by daclizumab, a CD25-blocking antibody that suppresses MS-associated inflammation. The objective of this study was to investigate the molecular mechanisms underlying the cytotoxicity of NK cells toward activated T cells. We demonstrated that NK cells induce caspase-independent apoptosis that requires NK cell degranulation and causes mitochondrial dysfunction in activated T cells. While both granzymes A (GrA) and K (GrK) can mediate this form of apoptosis, quantitatively we observed preferential transfer of GrK to target cells. Consequently, gene silencing of GrK in the NK-92 cell line, which retains functional characteristics of CD56bright NK cells, profoundly inhibited the ability of NK-92 to kill activated syngeneic T cells. Finally we demonstrated that daclizumab treatment significantly enhanced this newly defined mechanism of cytotoxicity by CD56bright NK cells. Our study represents the first example of the important physiological role GrK plays in immunoregulation of adaptive immunity in humans and indicates that therapeutic exploitation of this pathway is beneficial in controlling autoimmunity.

Multiple sclerosis (MS) is a chronic inflammatory disorder of the central nervous system (CNS). Neither the antigenic target(s) nor the cell population(s) responsible for CNS tissue destruction in MS have been fully defined. The objective of this study was to simultaneously determine the antigen (Ag)-specificity and phenotype of un-manipulated intrathecal CD4+ and CD8+ T cells of patients with relapsing-remitting and progressive MS compared to subjects with other inflammatory neurological diseases. We applied a novel Ag-recognition assay based on co-cultures of freshly obtained cerebrospinal fluid T cells and autologous dendritic cells pre-loaded with complex candidate Ag's. We observed comparably low T cell responses to complex auto-Ag's including human myelin, brain homogenate, and cell lysates of apoptotically modified oligodendroglial and neuronal cells in all cohorts and both compartments. Conversely, we detected a strong intrathecal enrichment of Epstein-Barr virus- and human herpes virus 6-specific (but not cytomegalovirus-specific) reactivities of the Th1-phenotype throughout all patients. Qualitatively, the intrathecal enrichment of herpes virus reactivities was more pronounced in MS patients. This enrichment was completely reversed by long-term treatment with the IL-2 modulating antibody daclizumab, which strongly inhibits MS disease activity. Finally, we observed a striking discrepancy between diminished intrathecal T cell proliferation and enhanced cytokine production of herpes virus-specific T cells among progressive MS patients, consistent with the phenotype of terminally differentiated cells. The data suggest that intrathecal administration of novel therapeutic agents targeting immune cells outside of the proliferation cycle may be necessary to effectively eliminate intrathecal inflammation in progressive MS.