Noniterative biexponential fluorescence lifetime imaging in the investigation of cellular metabolism by means of NAD(P)H autofluorescence

Abstract

The cofactors NADH and NADPH, hereafter NAD(P)H [NAD(P)= nicotinamide adenine dinucleotide (phosphate)], belong to the principal endogenous indicators of energetic cellular metabolism. Since the metabolic activity of cells is given by the ratio between the concentrations of free and protein-bound NAD(P)H, the development of autofluorescence techniques which accurately measure the modifications to this ratio is particularly significant. Hitherto the methods applied in the monitoring of cellular metabolism have provided either imprecise results, due to interference of the NAD(P)H signal by perturbing factors, or they have required a complicated internal calibration. We employ biexponential fluorescence lifetime imaging (FLIM) in order to discriminate between the free and protein-bound NAD(P)H without any previous calibration. Thus, we have obtained directly, and for the first time, a high-resolution map of cellular metabolism, that is, an image of the contribution of the protein-bound NAD(P)H to the cumulative NAD(P)H fluorescence signal. Moreover, we demonstrate that protein-NAD(P)H complexes characterised by different fluorescence lifetimes are not uniformly distributed all over the cell, as assumed until now, but are concentrated in certain cellular regions. The different fluorescence lifetimes indicate either different protein-NAD(P)H complexes or different bond strengths between NAD(P)H and the protein in these complexes. Since an important aspect in biological applications is to monitor the dynamics of the relevant processes (such as cellular metabolism), rapid dynamical techniques, for example, rapid biexponential fluorescence lifetime imaging, are needed. Furthermore, it is necessary to reduce the evaluation effort as much as possible. Most of the evaluation techniques in multiexponential FLIM are time-expensive iterative methods. The few exceptions are connected with a loss of information, for example, global analysis; or a loss in accuracy, for example, the rapid evaluation technique (RLD). We implement for the first time in FLIM a noniterative, nonrestrictive method originally developed by Prony for approximations of multiexponential decays. The accuracy of this method is verified in biexponential FLIM experiments in time-domain on mixtures of two chromophores both in homogenous and in heterogeneous media. The resulting fluorescence lifetimes agree (within error margins) with the lifetimes of the pure substances determined in monoexponential FLIM experiments. The rapidity of our evaluation method as compared to iterative pixel-by-pixel methods is evidenced by a reduction of the evaluation time by more than one order of magnitude. Furthermore, the applicability of this method for the biosciences is demonstrated in the investigation of cellular metabolism by means of NAD(P)H endogenous fluorescence.