Dynamic signatures of lipid droplets as new markers to quantify cellular metabolic changes

The metabolic properties of live cells are very susceptible to intra- or extra-cellular perturbations, making their measurements challenging tasks. We show that the dynamics of lipid droplets (LDs) carry information to measure the lipid metabolism of live cells. Coherent anti-Stokes Raman scattering microscopy was used to statistically quantify LD dynamics in living cells in a label-free manner. We introduce dynamic signatures of cells derived from the LD displacement, speed, travel length, and directionality, which allows for the detection of cellular changes induced by stimuli such as fluorescent labeling, temperature change, starvation, and chemical treatment. Histogram fittings of the dynamic signatures using lognormal distribution functions provide quantification of changes in cellular metabolic states. The LD dynamics also enable separation of subpopulations of LDs correlated with different functions. We demonstrate that LD dynamics are new markers to quantify the metabolic changes in live cells.