Application of copper (I) selective ligands for PET imaging of reactive oxygen species through metabolic trapping

Abstract

IntroductionReactive oxygen species (ROS) are attractive targets for clinical PET imaging. In this study, we hypothesized that PET imaging of ROS would be possible by using chelating ligands (L) that form stable complexes with copper (I) but not with copper (II), based on metabolic trapping. Namely, when [64Cu][CuI(L)2]+ is oxidized by ROS, the oxidized complex will release [64Cu]Cu2+. Then, the released [64Cu]Cu2+ will be trapped inside the cell, resulting in PET signal depending on the redox potential of ROS. To examine the potential of this novel molecular design for ROS imaging, we synthesized copper (I) complexes with bicinchoninic acid (BCA) disodium salt and bathocuproinedisulfonic acid (BCS) disodium salt and evaluated their reactivity with several kinds of ROS. In addition, the cellular uptake of [64Cu][CuI(BCS)2]3− and the stability of [64Cu][CuI(BCS)2]3− in a biological condition were also evaluated. Methods[64Cu]Cu2+ was reduced to [64Cu]Cu+ by ascorbic acid and coordinated with BCA and BCS in the acetate buffer to synthesize [64Cu][CuI(BCA)2]3− and [64Cu][CuI(BCS)2]3−. The radiochemical yields were determined by thin-layer chromatography (TLC). After [64Cu][CuI(BCS)2]3− was incubated with hydroxyl radical, lipid peroxide, superoxide, and hydrogen peroxide, the percentage of released [64Cu]Cu2+ from the parent complex was evaluated by TLC. HT-1080 human fibrosarcoma cells were treated with 0.1 % Dimethyl sulfoxide (control), imidazole ketone erastin (IKE), or IKE + ferrostatin-1 (Fer-1). Then, the uptake of [64Cu][CuI(BCS)2]3− to HT-1080 cells in each group was evaluated as %Dose/mg protein. Lastly, [64Cu][CuI(BCS)2]3− was incubated in human plasma, and its intact ratio was determined by TLC. ResultsThe radiochemical yield of [64Cu][CuI(BCS)2]3− (86 ± 1 %) was higher than that of [64Cu][CuI(BCA)2]3− (44 ± 3 %). [64Cu][CuI(BCA)2]3− was unstable and partially decomposed on TLC. After [64Cu][CuI(BCS)2]3− was reacted with hydroxyl radical, lipid peroxide, and superoxide, 67 ± 2 %, 44 ± 13 %, and 22 ± 3 % of total radioactivity was detected as [64Cu]Cu2+, respectively. On the other hand, the reaction with hydrogen peroxide did not significantly increase the ratio of [64Cu]Cu2+ (4 ± 1 %). These results suggest that [64Cu][CuI(BCS)2]3− could be used for detecting high-redox-potential ROS such as hydroxyl radical and lipid peroxide with high selectivity. The cellular uptake values of [64Cu][CuI(BCS)2]3− in the control, IKE, and Fer-1 group were 42 ± 2, 54 ± 2, and 47 ± 5 %Dose/mg protein (n = 3), respectively, suggesting the ROS specific uptake of [64Cu][CuI(BCS)2]3−. On the other hand, the intact ratio after the incubation of [64Cu][CuI(BCS)2]3− in human plasma was 9 ± 5 %. ConclusionPET imaging of ROS would be possible by using a copper (I) selective ligand, based on metabolic trapping. Although improvement of the membrane permeability and the stability of copper (I) complexes is required, the present results pave the way for the development of novel 64Cu-labeled complexes for PET imaging of ROS.