Cheng et al. reply

Replying to M. Schwarzländer et al. Nature 514, 10.1038/nature13858 (2014) In the accompanying Comment1, Schwarzländer et al. challenged our recent study2 because they failed to reproduce our previous finding that the fluorescence intensity of purified circularly permuted yellow fluorescent protein (cpYFP) increases in response to oxygen and superoxide anions produced by xanthine (X) plus xanthine oxidase (XO)3. Starting from a ‘fully reduced’ state (incubation with 10 mM dithiothreitol for >3 h) and in the presence of 75 mM HEPES, we demonstrated that cpYFP exhibits a twofold fluorescence increase after oxygenation, and an additional twofold increase after the subsequent addition of X plus XO, which could not be accounted for by solvent (potassium hydroxide)-induced alkalization. Furthermore, the xanthine plus xanthine oxidase-induced increase in cpYFP fluorescence was reversed by Cu/Zn superoxide dismutase (600 U ml−1). We also found that the fluorescence intensity of fully reduced cpYFP increased >fourfold after incubation with 1 mM aldrithiol. Notably, recombinant cpYFP purified in the absence of dithiothreitol treatment exhibits a high fluorescence comparable to that of the fully oxidized state, indicating the high susceptibility of cpYFP to oxidation in non-reducing environments3. Therefore, ensuring a fully reduced state of cpYFP is essential for the probe to sense superoxide in vitro. This property is probably the reason that the probe functions readily as a reversible superoxide biosensor when targeted to the reduced environment of the mitochondrial matrix. Unfortunately, from the brief description of the methods and limited data provided by Schwarzländer et al.1, it is not possible to determine whether cpYFP was fully reduced in their experiments, or whether sufficient precautions were taken to prevent oxidation of the probe. Moreover, in our experiments cpYFP was expressed in Escherichia coli BL21(DE3)LysS cells, whereas Schwarzländer et al.1 used E. coli Origami, a trxB (thioredoxin reductase) mutant strain that also lacks glutathione reductase needed to fully limit cysteine oxidation4, which could result in an increased oxidative status of their purified cpYFP rendering it non-responsive to superoxide.