ABSTRACT
R-2-hydroxyglutarate accumulates to millimolar levels in cancer cells with gain-of-function isocitrate dehydrogenase 1/2 mutations. These levels of R-2-hydroxyglutarate affect 2-oxoglutarate-dependent dioxygenases. Both metabolite enantiomers, R- and S-2-hydroxyglutarate, are detectible in healthy individuals, yet their physiological function remains elusive. Here we show that 2-hydroxyglutarate accumulates in mouse CD8+ T cells in response to T-cell receptor triggering, and accumulates to millimolar levels in physiological oxygen conditions through a hypoxia-inducible factor 1-alpha (HIF-1α)-dependent mechanism. S-2-hydroxyglutarate predominates over R-2-hydroxyglutarate in activated T cells, and we demonstrate alterations in markers of CD8+ T-cell differentiation in response to this metabolite. Modulation of histone and DNA demethylation, as well as HIF-1α stability, mediate these effects. S-2-hydroxyglutarate treatment greatly enhances the in vivo proliferation, persistence and anti-tumour capacity of adoptively transferred CD8+ T cells. Thus, S-2-hydroxyglutarate acts as an immunometabolite that links environmental context, through a metabolic-epigenetic axis, to immune fate and function.
Subject(s)
CD8-Positive T-Lymphocytes/cytology , CD8-Positive T-Lymphocytes/drug effects , Cell Differentiation/drug effects , Glutarates/pharmacology , Animals , CD8-Positive T-Lymphocytes/immunology , CD8-Positive T-Lymphocytes/metabolism , DNA/chemistry , DNA/metabolism , DNA Methylation/drug effects , Dioxygenases/metabolism , Glutarates/immunology , Glutarates/metabolism , Histones/metabolism , Homeostasis/drug effects , Hypoxia/metabolism , Hypoxia-Inducible Factor 1, alpha Subunit/metabolism , Ketoglutaric Acids/metabolism , Lymphocyte Activation , Lysine/metabolism , Mice , Oxygen/metabolism , Protein Stability , Receptors, Antigen, T-Cell/immunology , Von Hippel-Lindau Tumor Suppressor Protein/metabolismABSTRACT
The nonlinear discriminant function obtained using a minimum squared error cost function can be shown to be directly related to the nonlinear Fisher discriminant (NFD). With the squared error cost function, the orthogonal least squares (OLS) algorithm can be used to find a parsimonious description of the nonlinear discriminant function. Two simple classification techniques will be introduced and tested on a number of real and artificial data sets. The results show that the new classification technique can often perform favourably compared with other state of the art classification techniques.