Quenching of superoxide radicals by green fluorescent protein

Abstract

Green fluorescent proteins (GFP) are widely used in vivo molecular markers. These proteins are particularly resistant, and maintain function, under a variety of cellular conditions such as pH extremes and elevated temperatures. Green fluorescent proteins are also abundant in several groups of marine invertebrates including reef-forming corals. While molecular oxygen is required for the post-translational maturation of the protein, mature GFPs are found in corals where hyperoxia and reactive oxygen species (ROS) occur due to the photosynthetic activity of algal symbionts. In vitro spin trapping electron paramagnetic resonance and spectrophotometric assays of superoxide dismutase (SOD)-like enzyme activity show that wild type GFP from the hydromedusa, Aequorea victoria, quenches superoxide radicals (O2*-)) and exhibits SOD-like activity by competing with cytochrome c for reaction with O2*-. When exposed to high amounts of O2*- the SOD-like activity and protein structure of GFP are altered without significant changes to the fluorescent properties of the protein. Because of the distribution of fluorescent proteins in both the epithelial and gastrodermal cells of reef-forming corals we propose that GFP, and possibly other fluorescent proteins, can provide supplementary antioxidant protection.

Figure 1

a) EPR of EMPO-OOH adduct in buffer alone as a control (A), in the presence of egg albumin (14 μM) (B), and in the presence of 13.8 μM GFP (C). b) EPR of DEPMPO-OOH adduct in the absence (A) and presence (B) of GFP (18.5 μM). O₂^•− were generated by the hypoxanthine/xanthine oxidase system in 50 mM phosphate buffer at pH 7.4, 0.5 mM DTPA and 50 mM EMPO or DEPMPO.

Figure 2

Reduction of cytochrome c by O₂^•− generated using the hypoxanthine/xanthine oxidase system and inhibition by SOD or GFP. Conditions: 25 μM cytochrome c, 0.1 mM EDTA, 1300 U ml⁻¹ catalase, 6 uM hypoxanthine per injection (stars denote injection times), and 0.1 U ml⁻¹ xanthine oxidase. Kinetics were performed in the presence of either egg albumin (1.5 uM), GFP (1.5 μM) or Cu/Zn-SOD (1.5 U ml⁻¹, 0.015 μM) from bovine erythrocytes (Sigma, EC 1.15.1.1) Inset: Superoxide quenching activity of Aequorea victoria GFP at increasing protein concentrations (2.4 to 61.7 nM) and a constant 100 nM (O₂^•−).

Figure 3

Fluorescence excitation/emission spectra of GFP before (dotted line) and after (black line) exposure to O₂^•−. Conditions: GFP (0.37 μM) in 50 mM Tris-HCl buffer, pH 8.0, with or without 0.2 mM hypoxanthine, and 0.1 U ml⁻¹ xanthine oxidase.

Figure 4

a) Electropherograms of GFP (0.44 μg μl⁻¹) in the presence and absence of O₂^•− (A) Buffer control, (B) GFP alone in buffer, (C) GFP + O₂^•−. Conditions: 0.1 M phosphate buffer, pH 7.2, 0.4 mM DTPA, 0.17 mM hypoxanthine and 0.1 U ml⁻¹ xanthine oxidase. b) Electropherograms of GFP (0.27 μg μl⁻¹) under denaturing conditions. (A) Buffer control, (B) GFP alone in buffer, (C) GFP + O₂^•− + mannitol, (D) GFP + O₂^•− but no mannitol. Arrows denote the conversion of species (1) to (2) following O₂^•− addition. Conditions: 0.1 M phosphate buffer, pH 7.2, 50 mM mannitol, 0.1 mM hypoxanthine and 0.1 U ml⁻¹ xanthine oxidase. In order to correct for the run-to-run variations, the data were normalized by dividing the migration times for each sample by the migration time of the phosphate buffer peak in native CE or by that of Orange G in SDS-CGE to give a relative migration time (RMT).

Figure 4

a) Electropherograms of GFP (0.44 μg μl⁻¹) in the presence and absence of O₂^•− (A) Buffer control, (B) GFP alone in buffer, (C) GFP + O₂^•−. Conditions: 0.1 M phosphate buffer, pH 7.2, 0.4 mM DTPA, 0.17 mM hypoxanthine and 0.1 U ml⁻¹ xanthine oxidase. b) Electropherograms of GFP (0.27 μg μl⁻¹) under denaturing conditions. (A) Buffer control, (B) GFP alone in buffer, (C) GFP + O₂^•− + mannitol, (D) GFP + O₂^•− but no mannitol. Arrows denote the conversion of species (1) to (2) following O₂^•− addition. Conditions: 0.1 M phosphate buffer, pH 7.2, 50 mM mannitol, 0.1 mM hypoxanthine and 0.1 U ml⁻¹ xanthine oxidase. In order to correct for the run-to-run variations, the data were normalized by dividing the migration times for each sample by the migration time of the phosphate buffer peak in native CE or by that of Orange G in SDS-CGE to give a relative migration time (RMT).