The relationship between increased levels of RAGE ligands and increased RAGE expression in activated T cells suggests a ‘‘feed forward’’ mechanism that may explain the common finding of increased RAGE expression on T cells from patients with T1 and T2D. A limitation of our studies is that since RAGE undergoes a variety of post-translational modifications, it is possible that an isoform of the molecule is expressed by activated T cells that is not identified by the antibody we used for detection. In tissues, the full length form of RAGE is found most frequently, but RAGE undergoes a variety of splice events resulting, most commonly, in the production of a secreted form of RAGE from a frameshift at the C terminus which removes transmembrane and cytoplasmic domains. In pathologic states the level of expression of the splice variant may change which results in increased levels of sRAGE in plasma. Nonetheless, our studies of unactivated cells indicate that there are clear differences in the patterns of RAGE expression on T cells from healthy control subjects and patients with diabetes. RAGE appears to modulate the phenotype of CD4+ T cells. In mice, we found reduced expression of IFNc by RAGE deficient cells and our studies with human cells show increased expression of IL-17 and CD107a in RAGE+ T cells. Signaling through RAGE has been shown to involve interactions between the FH1 domain of mammalian Diaphanous-1 that interacts with the cytoplasmic tail of RAGE and induces several intermediaries including NF-kB, MAPKs, PI3K/Akt, Rho GTPases, Jak/STAT, and Src family kinases. Modulation of TCR signaling via these intermediates may affect the phenotype of activated cells. This observation may help to explain the observations of others concerning increased IL-17 production in responses to antigen in patients with T1DM. Nakamura et al found that circulating AGEs and sRAGE are independent determinants of serum monocyte chemoattractant protein-1 levels in patients with type 2 diabetes suggesting a direct relationship between immune cell activation and AGE levels. However, since RAGE expression on T cells is affected by both the presence of ligands as well as TCR stimulation, our findings suggest a more complex control of RAGE expression than simply the availability of ligands. In addition, the increased proportion of RAGE+ CD4+ T cells in cultures with peptide presented by Class I MHC molecules raises the possibility that RAGE may be induced on neighboring T cells possibly by cell:cell contact or by soluble factors that have not been identified. Although RAGE expression is associated with hyperglycemia rather than the autoimmune process that causes T1D, our findings may have relevance for understanding the more rapid tempo of the disease that is seen after the development of hyperglycemia. Beta cell destruction in T1DM is believed to progress in a linear fashion from the first appearance of autoantibodies until complete elimination of b cells, the tempo of the disease increases once hyperglycemia develops but studies have shown a 10 fold greater loss of insulin secretion after development of hyperglycemia compared to before diagnosis and analysis of changes in C-peptide secretion in the DCCT suggested that tight glycemic control was associated with reduced decline in b cell function. In summary, we have described, for the first time, the expression of RAGE in human T cells, and have explored factors that control its expression, and the relationship between RAGE expression and T cell function.