Type 1 diabetes is an autoimmune disease typically characterized by a long preclinical stage in which insulin-producing β cells in the pancreatic islets are destroyed by autoimmune T cells. In the June issue of Immunity, Lewis Lanier and colleagues use the non-obese diabetic (NOD) mouse to show that the ligand retinoic acid early inducible (RAE)-1, which is present in pre-diabetic islets, activates the receptor NKG2D, which is expressed on the surface of a subset of T cells and natural killer (NK) cells; furthermore, this interaction is essential for disease progression. When treated during the pre-diabetic stage, a non-depleting anti-NKG2D antibody completely prevented disease.

The NOD mouse is widely used as a model of human type 1 diabetes. Usually, inflammation of the islets is observed 10–20 weeks before T-cell destruction of the insulin-producing cells; this process and the time delay is not well understood. NKG2D is a costimulatory receptor expressed on activated NK cells and certain groups of T cells, and there is evidence that it might carry out surveillance against cancer and infections. The ligands of NKG2D, including RAE-1, are known to be molecules whose expression is triggered by oncological transformation and infection with viral and bacterial pathogens. Costimulatory molecules on T cells and other immune cells are crucial for the autoimmune response. Expression of RAE-1 family members is strictly regulated in normal cells, and little is expressed on healthy adult tissues. However, the authors found that RAE-1 was expressed on islet cells in NOD mice, but not in other strains such as C57BL/6 or BALB/c. Expression was also detected in the pancreases of adult immune-deficient NOD mice (which lack T cells and do not develop diabetes), indicating that RAE-1 expression is independent of ongoing autoimmune response.

To demonstrate the role of NKG2D in the progression of diabetes, the authors transferred T cells from diabetic NOD mice into immune-deficient NOD mice. Diabetes developed in all control mice, but not those treated with anti-NKG2D antibodies. The authors showed that the therapy impaired the clonal expansion and function of autoreactive T cells in the pancreas, although it did not prevent the initial activation of the T cells.

These findings are important because unlike other treatments reported in the NOD mouse this therapy is effective when administered relatively late in disease progression. Interestingly, NKG2D has been detected on T cells — isolated from the synovial tissue of rheumatoid arthritis patients — in the presence of one of its ligands, which raises the posibility that this receptor could play a detrimental role in other autoimmune diseases.