Insights into NAD homeostasis in the compartmentalized cell

Abstract

The compartmentalized nature of eukaryotic cells requires the distribution of enzymes, metabolites, and cofactors among the organelles. This includes nicotinamide adenine dinucleotide (NAD), an essential coenzyme, precursor, and substrate to many cellular reactions. For instance, this dinucleotide is used by the NAD+-dependent deacetylase sirtuin2 (SIRT2). The cytosolic location of this protein is in apparent contradiction to the nuclear location of some of its targets. In this study, whether a hitherto unidentified nuclear SIRT2 isoform could account for nuclear SIRT2-mediated deacetylation was investigated. A novel human SIRT2 isoform (SIRT2 isoform 5) was identified and shown to reside in the nucleus. Strikingly, this protein did not exhibit deacetylase activity towards several known SIRT2 targets. However, it retained the ability to interact with p300 in a NAD+-dependent manner. These results suggest a non-catalytic function for SIRT2-isoform 5 in the nucleus of human cells. The distribution of NAD+-consuming enzymes in subcellular compartments highlights the necessity for the maintenance of the different NAD+ pools. Mitochondria contain a substantial amount of the total cellular NAD+. In yeast and plants, mitochondrial NAD+ transporters ensure the maintenance of that pool, whereas in mammals such a transporter could not be identified so far. To address this problem, the closest human homologs to the Arabidopsis thaliana NAD+ transporter NDT2 were expressed in human cells. None of them increased mitochondrial NAD+ availability, indicating they do not function as NAD+ carriers in humans. This finding argues for autonomous NAD biosynthesis in human mitochondria. Indeed, nicotinamide mononucleotide adenylyltransferase 3 (NMNAT3), an isoform of the enzyme that catalyzes the last step of NAD synthesis, was reported to be localized to the mitochondria. However, all previous studies were done with a recombinant protein, and the presence of the endogenous protein is still debated. This study conclusively demonstrates the presence of the endogenous NMNAT3 protein in human mitochondria, supporting the idea of autonomous biosynthesis within these organelles. On the basis of the observation that the expression of plant and yeast NAD+ transporters in human cells strongly increased mitochondrial NAD+ content, the consequences of increased mitochondrial NAD+ availability were investigated. Stable expression of the plant NAD+ transporter NDT2 in HEK293 cells resulted in growth retardation as well as a metabolic shift from mitochondrial respiration to glycolysis, and conferred increased resistance towards cell death induced by NAD+-depletion. These results suggest that distribution of NAD+ between the cytosol and the mitochondria is a major determinant of the metabolic profile of human cells.