Nonlinear Regression Models for Determination of Nicotinamide Adenine Dinucleotide Content in Human Embryonic Stem Cells

Abstract

Recent evidence suggests that energy metabolism contributes to molecular mechanisms controlling stem cell identity. For example, human embryonic stem cells (hESCs) receive their metabolic energy mostly via glycolysis rather than mitochondrial oxidative phosphorylation. This suggests a connection of metabolic homeostasis to stemness. Nicotinamide adenine dinucleotide (NAD) is an important cellular redox carrier and a cofactor for various metabolic pathways, including glycolysis. Therefore, accurate determination of NAD cellular levels and dynamics is of growing importance for understanding the physiology of stem cells. Conventional analytic methods for the determination of metabolite levels rely on linear calibration curves. However, in actual practice many two-enzyme cycling assays, such as the assay systems used in this work, display prominently nonlinear behavior. Here we present a diaphorase/lactate dehydrogenase NAD cycling assay optimized for hESCs, together with a mechanism-based, nonlinear regression models for the determination of NAD(+), NADH, and total NAD. We also present experimental data on metabolic homeostasis of hESC under various physiological conditions. We show that NAD(+)/NADH ratio varies considerably with time in culture after routine change of medium, while the total NAD content undergoes relatively minor changes. In addition, we show that the NAD(+)/NADH ratio, as well as the total NAD levels, vary between stem cells and their differentiated counterparts. Importantly, the NAD(+)/NADH ratio was found to be substantially higher in hESC-derived fibroblasts versus hESCs. Overall, our nonlinear mathematical model is applicable to other enzymatic amplification systems.