Insight into redox regulation of apoptosis in cancer cells with multiparametric live-cell microscopy

Abstract

Cellular redox status and the level of reactive oxygen species (ROS) are important regulators of apoptotic potential, playing a crucial role in the growth of cancer cell and their resistance to apoptosis. However, the relationships between the redox status and ROS production during apoptosis remain poorly explored. In this study, we present an investigation on the correlations between the production of ROS, the redox ratio FAD/NAD(P)H, the proportions of the reduced nicotinamide cofactors NADH and NADPH, and caspase-3 activity in cancer cells at the level of individual cells. Two-photon excitation fluorescence lifetime imaging microscopy (FLIM) was applied to monitor simultaneously apoptosis using the genetically encoded sensor of caspase-3, mKate2-DEVD-iRFP, and the autofluorescence of redox cofactors in colorectal cancer cells upon stimulation of apoptosis with staurosporine, cisplatin or hydrogen peroxide. We found that, irrespective of the apoptotic stimulus used, ROS accumulation correlated well with both the elevated pool of mitochondrial, enzyme-bound NADH and caspase-3 activation. Meanwhile, a shift in the contribution of bound NADH could develop independently of the apoptosis, and this was observed in the case of cisplatin. An increase in the proportion of bound NADPH was detected only in staurosporine-treated cells, this likely being associated with a high level of ROS production and their resulting detoxification. The results of the study favor the discovery of new therapeutic strategies based on manipulation of the cellular redox balance, which could help improve the anti-tumor activity of drugs and overcome apoptotic resistance.

Figure 1

Caspase-3 activity in CT26 cancer cells stably expressing the genetically encoded FRET-based sensor, mKate2-DEVD-iRFP, upon induction of apoptosis. (A) Time-lapse FLIM images of the donor mKate2 before (0 h) and after treatment with staurosporine (STS), cisplatin or hydrogen peroxide. Image size is 213 × 213 μm. Scale bar: 50 μm. (B) Fluorescence lifetime of mKate2 in the individual cells shown in (A). Individual cells in (A) are numbered.

Figure 2

Optical redox ratio FAD/NADH in CT26 cancer cells stably expressing the genetically encoded FRET-based sensor, mKate2-DEVD-iRFP, upon induction of apoptosis. (A) Time-lapse imaging of the ratio FAD/NADH before (0 h) and after treatment with staurosporine (STS), cisplatin or hydrogen peroxide. Image size is 213 × 213 μm. Scale bar: 50 μm. (B) Quantification of the optical redox ratio FAD/NADH in the cells. Mean ± SD, n = 20–50 cells. *p ≤ 0.05 with control (0 h). Images were acquired from the same fields of view as in Fig. 1.

Figure 3

FLIM of NAD(P)H in CT26 cancer cells stably expressing the genetically encoded FRET-based sensor, mKate2-DEVD-iRFP, upon induction of apoptosis. (A) Time-lapse images of the relative contributions of bound NADH (a₂) and bound NADPH (a₃) before (0 h) and after treatment with staurosporine (STS), cisplatin or hydrogen peroxide. Image size is 213 × 213 μm. Scale bar: 50 μm. (B) Quantification of the relative contributions of the bound NADH (a₂) and bound NADPH (a₃) in the cells. Mean ± SD, n = 20–50 cells. *p ≤ 0.05 with control (0 h). NAD(P)H fluorescence decay curves for STS were processed with three-exponential fitting; and for cisplatin and hydrogen peroxide—with bi-exponential fitting. Images were acquired from the same fields of view as in Figs. 1 and 2.

Figure 4

Simultaneous in vivo FLIM of mKate2 and NAD(P)H in CT26 tumors expressing the genetically encoded FRET-based sensor of caspase-3 activity, mKate2-DEVD-iRFP. (A) Representative FLIM images of mKate2 and NAD(P)H before (control) and after treatment with staurosporine (STS). Image size is 213 × 213 μm. Scale bar: 100 μm. (B) Quantification of the fluorescence lifetime of mKate2 and the fluorescence lifetime of bound NADH (τ₂) in the tumor cells. Mean ± SD, n = 20–50 cells. *p ≤ 0.00001 with control. The NAD(P)H fluorescence decay curves were processed with bi-exponential fitting. (C) Histopathological analysis of the control and STS-treated tumors. H&E staining. Scale bar: 50 μm.

Figure 5

ROS level in CT26 cancer cells stably expressing the genetically encoded FRET-based sensor, mKate2-DEVD-iRFP, upon induction of apoptosis. Left: Time-lapse microscopic images of cells stained with the ROS-sensitive probe H₂DCFDA, before (0 h) and after treatment with staurosporine (STS), cisplatin or hydrogen peroxide. Image size is 213 × 213 μm. Scale bar: 50 μm. Right: Quantification of fluorescence intensity of DCF in the cells. Box-and-Whisker plots display the median, 25th and 75th percentiles, minimum and maximum. n = 20–50 cells. *p ≤ 0.05 with control (0 h).

Figure 6

Correlations between caspase-3 activity (τ_m mKate2), the protein-bound NADH fraction (a₂-NADH) and the ROS levels (DCF) in cancer cells upon induction of apoptosis. Blue dots are the measurements for individual cells. Pearson’s correlation coefficient and p-value are indicated on each plot. The solid line represents the regression line.

Figure 7

Schematic of the associations between the redox ratio FAD/NADH, the metabolic cofactors NADH and NADPH, the production of ROS, and caspase-3 activity in cancer cells. Blue arrows show the order of events. Gray dashed arrows show correlations between the events revealed for the specific apoptotic stimuli (STS, cisplatin or H₂O₂).