Single Cell Imaging of Metabolism with Fluorescent Biosensors

Abstract

Cells utilize various signal transduction networks to regulate metabolism. Nevertheless, a quantitative understanding of the relationship between growth factor signaling and metabolic state at the single cell level has been lacking. The signal transduction and metabolic states could vary widely among individual cells. However, such cell-to-cell variation might be masked by the bulk measurements obtained from conventional biochemical methods. To assess the spatiotemporal dynamics of metabolism in individual intact cells, we developed genetically encoded biosensors based on fluorescent proteins. As a key redox cofactor in metabolism, NADH has been implicated in the Warburg effect, the abnormal metabolism of glucose that is a hallmark of cancer cells. To date, however, sensitive and specific detection of NADH in the cytosol of individual live cells has been difficult. We engineered a fluorescent biosensor of NADH by combining a circularly permuted green fluorescent protein variant with a bacterial NADH-binding protein Rex. The optimized biosensor Peredox reports cytosolic \(NADH:NAD^+\) ratios in individual live cells and can be calibrated with exogenous lactate and pyruvate. Notably pH resistant, this biosensor can be used in several cultured and primary cell types and in a high-content imaging format. We then examined the single cell dynamics of glycolysis and energy-sensing signaling pathways using Peredox and other fluorescent biosensors: AMPKAR, a sensor of the AMPK activity; and FOXO3-FP, a fluorescently-tagged protein domain from Forkhead transcription factor FOXO3 to report on the PI3K/Akt pathway activity. With perturbation to growth factor signaling, we observed a transient response in the cytosolic \(NADH:NAD^+\) redox state. In contrast, with partial inhibition of glycolysis by iodoacetate, individual cells varied substantially in their responses, and cytosolic \(NADH:NAD^+\) ratios oscillated between high and low states with a regular, approximately half-hour period, persisting for hours. These glycolytic NADH oscillations appeared to be cell-autonomous and coincided with the activation of the PI3K/Akt pathway but not the AMPK pathway. These results suggest a dynamic coupling between growth factor signaling and metabolic parameters. Overall, this thesis presents novel optical tools to assess metabolic dynamics – and to unravel the elaborate and complex integration of glucose metabolism and signaling pathways at the single cell level.