Separating NADH and NADPH fluorescence in live cells and tissues using FLIM

Abstract

NAD is a key determinant of cellular energy metabolism. In contrast, its phosphorylated form, NADP, plays a central role in biosynthetic pathways and antioxidant defence. The reduced forms of both pyridine nucleotides are fluorescent in living cells but they cannot be distinguished, as they are spectrally identical. Here, using genetic and pharmacological approaches to perturb NAD(P)H metabolism, we find that fluorescence lifetime imaging (FLIM) differentiates quantitatively between the two cofactors. Systematic manipulations to change the balance between oxidative and glycolytic metabolism suggest that these states do not directly impact NAD(P)H fluorescence decay rates. The lifetime changes observed in cancers thus likely reflect shifts in the NADPH/NADH balance. Using a mathematical model, we use these experimental data to quantify the relative levels of NADH and NADPH in different cell types of a complex tissue, the mammalian cochlea. This reveals NADPH-enriched populations of cells, raising questions about their distinct metabolic roles.

Figure 1. τ_bound reflects the enzyme-bound NADPH/NADH ratio in intact cells.

(a,b) A biexponential decay model adequately described the NAD(P)H fluorescence decay measured in wtHEK293 cells (IRF, instrument response function). The mean was 1.24±0.08 compared with 4±1 with a monoexponential fit (representative data from n=17 experiments). (c,d) Representative colour-coded images and mean τ_bound and α_bound values in NADK+ and NADK− HEK293 cells prior and following treatment with EGCG (100 μM), a competitive inhibitor of NADPH binding. Scale bar, 20 μm. Error bars indicate±s.d., *P<0.05 (two-tailed Student’s t-test, n=9).

Figure 2. NAD(P)H fluorescence decay responses to metabolic perturbation are mechanism dependent.

(a–d) Time series of NAD(P)H fluorescence intensity following treatments chosen to perturb oxidative (a: respiratory chain inhibition by rotenone, b: uncoupling with FCCP) and glycolytic metabolism (glucose replaced by deoxyglucose with c: pyruvate or d: lactate supplied as substrate). (e) Colour-coded images and (f,g) quantification of changes in τ_bound and α_bound on application of treatment. Scale bar, 20 μm. Error bars indicate±s.d., *P<0.05 (two-tailed Student’s t-test, n=9).

Figure 3. Supporting cells in the mammalian cochlea exhibit increased enzyme-bound NADPH.

(a) Mean τ_bound and (b) α_bound values in outer hair cells (OHC’s) and adjacent outer pillar ‘supporting’ cells (OPC’s) under control conditions and following application of EGCG (200 μM). Error bars indicate±s.d., *P<0.05 (two-tailed Student’s t-test, n=11). (c,d) Corresponding representative FLIM images colour coded for the mean parameter value in each cell. Scale bar, 25 μm. (e) Schematic diagram showing organ of Corti in the cochlear explants, indicating the positions of OHC’s and OPC’s. (f) Mean fluorescence intensity in OPC’s and OHC’s. Error bars indicate±s.d., *P<0.05 (Wilcoxon signed-rank test, n=17). (g) Representative image following MCB staining for GSH concentration.