Study of Metabolite Repair in Eukaryotic Cells: Metabolic origin and fate of D-2-hydroxyglutarate in yeast and effect of NAD(P)HX repair deficiency on yeast and human cells

Abnormal metabolites, which are useless and can even be toxic, are constantly generated inside the cell by unwanted chemical reactions or by enzymatic side reactions. Metabolite repair enzymes clean the metabolite pool from these molecules. The proportion of proteins annotated as metabolite repair enzymes is currently very small but accumulating evidence suggests that a bigger part might be hidden among proteins of unknown function. The aim of this thesis was to study two of these metabolite repair systems and their physiological relevance in more detail as their importance is well illustrated through implication in disease processes. D-2-hydroxyglutaric aciduria, a severe human neurometabolic disorder, can be caused by a deficiency in the metabolite repair enzyme D-2-hydroxyglutarate (D-2HG) dehydrogenase. Higher levels of D-2HG have also been observed in cancerous cells with a mutated form of isocitrate dehydrogenase. Strikingly, in the model organism Saccharomyces cerevisiae, 2-hydroxyglutarate metabolism had remained completely unexplored. We elucidated the metabolic pathways involved in D-2HG formation and degradation in yeast using bioinformatics, metabolomics, yeast genetics, and classical biochemical tools. We discovered that Dld3, currently annotated as a D-lactate dehydrogenase, actually degrades D-2HG to α-ketoglutarate while reducing pyruvate to D-lactate, thereby acting as a transhydrogenase. We also demonstrated that the yeast phosphoglycerate dehydrogenases Ser3 and Ser33 are major sources for D-2HG formation. These findings paved the way to integrate 2HG and its associated genes into the yeast metabolic network and might help, on the long-term, to better understand underlying mechanisms in human disease as well. Other recently identified metabolite repair enzymes, NAD(P)HX dehydratase and NAD(P)HX epimerase (encoded in yeast by the YKL151C and YNL200C genes, respectively), specifically act on NADHX and NADPHX, hydrated and inactive forms of the central NADH and NADPH cofactors. Although extensively biochemically characterized, the physiological importance of these two enzymes still remains largely unclear. Only very recently, case reports were published indicating a correlation between NAD(P)HX repair deficiency and severe neuropathological symptoms starting in early childhood upon events of febrile illnesses and rapidly leading to a fatal outcome. We systematically analyzed extracts of NAD(P)HX repair deficient yeast and human cells using HPLC and LC-MS/MS methods. This enabled us to demonstrate that NADHX and NADPHX can be formed intracellularly. In the yeast system, NADHX accumulation, which could be modulated by the cultivation temperature, was accompanied by a decrease in intracellular NAD+ levels. Furthermore, we showed that NADHX interferes with serine metabolism by inhibiting the first step of the main synthesis pathway of this amino acid. In the human cell system, NAD(P)HX dehydratase deficiency led, as in yeast, to intracellular NADHX accumulation, but also to a marked decrease in cell viability after prolonged cultivation times. This is, to our knowledge, the first report about the effect of NADHX accumulation on cellular metabolism. Expanding our experimental strategy of combined transcriptomics and metabolomics approaches to the human cell model might ultimately lead to the discovery of the disease-causing cellular process. The findings in both projects led to an unexpected connection between NAD(P)HX and 2HG metabolism via the yeast homologues of 3-phoshpoglycerate dehydrogenase, Ser3 and Ser33. Both proteins catalyze the oxidation of 3-phosphoglycerate to 3-phosphohydroxypyruvate in the initial step of de novo serine biosynthesis with a concomitant reduction of α-ketoglutarate to D-2-hydroxyglutarate. By acting as transhydrogenases, they substantially, even though not exclusively, contribute to D-2HG formation in yeast. The very same enzymes were strongly inhibited in vitro and, as suggested by our findings, also in vivo by the presence of NADHX, leading to serine depletion in NAD(P)HX repair deficient cells