Positron emission tomography detection of human endothelial cell and fibroblast monolayers: effect of pretreament and cell density on 18FDG uptake

In this study, human umbilical vein endothelial cells (HUVEC) and human fibroblasts were cultured at different densities and PET was used to non-destructively monitor their glycolytic activity by measuring 18F-fluorodeoxyglucose (18FDG) uptake. Various cell preconditioning protocols were investigated by adjusting the following parameters to optimize 18FDG uptake: glucose starvation, insulin stimulation, glucose concentration, 18FDG incubation time, cell density and radiotracer efflux prevention.The conditions yielding optimal 18FDG uptake, and hence best detection sensitivity by PET, were as follows: 2-hour cell preconditioning by glucose deprivation with 1-hour insulin stimulation, followed by 1-hour 18FDG incubation and 15-minute stabilization in standard culture medium, prior to rinsing and PET scanning.A step-wise dependence of 18FDG uptake on glucose concentration was found, but a linear correlation between PET signal and cell density was observed. Detection thresholds of 36 ± 7 and 21 ± 4 cells were estimated for endothelial cells and fibroblasts, respectively.The growth of tissue substitutes is a dynamic process, requiring close monitoring of cell viability and function over time. Unfortunately, most of the commonly available cell assays (e.g., histology, immunofluorescence or immunocytochemistry) are time consuming and require sacrificing the culture. Although these techniques provide important information on cell phenotype and function, they are often only representatives of specific time points and selected samples within the culture. It remains difficult to trace the evolution of the cell or tissue cultures over time. Our inability to obtain continuous direct information on culture conditions and cells state in thick (few mm to cm range) 3D culture chambers represents a major weakness in understanding bioreactors performance. The main challenge nowadays is to find suitable real-time, non-invasive and, most importantly, non-destructive characterization met