Redox Reactions of Sanazole (AK- 2123) in Aqueous Solutions: A Pulse Radiolysis Study

SUDHIR KAPOOR 2 ROBIN MATHEW 1 NAGARAJ G. HUILGOL 1 TSUTOMU V. KAGIYA 0 CHERUPALLY KRISHNAN K. NAIR 3 0 Health Research Foundation , Kyoto, Japan 1 Division of Radiation Oncology, Nanavati Hospital , Vile parle, Mumbai 400 056, India 2 Radiation Chemistry & Chemical Dynamics Division 3 Radiation Biology Division, Bhabha Atomic Research Centre , Mumbai 400085, India (Received, July 21, 2000) (Revision received, October 23, 2000) (Accepted, November 8, 2000) Pulse radiolysis/Electron transfer/Reduction potential/Free radicals/Thiol depletion The redox chemistry of sanazole, an efficient hypoxic cell radiosensitizer, generally referred to as AK-2123, was studied by pulse radiolysis with eaq-, CO2-., 2-propanol radicals and .CH2OH radicals. AK-2123 reacts with eaq-, CO2-. and 2-propanol radicals at almost diffusion-controlled rates, producing a nitro radical anion (max = 290 nm) within a few microseconds. The decay kinetics of the radical anion was independent of the pH. The radical anion reacts with oxygen with a rate constant of 3.4 106 dm3 mol-1 s-1. An electron-transfer reaction was observed from the thymine radical anion to AK-2123. From redox equilibria with methyl viologen, the one-electron reduction potential of AK-2123 in aqueous solution, determined by pulse radiolysis, was estimated to be -0.33 0.02 V vs. NHE. Depletion of intracellular nonprotein thiols did not mitigate the radiosensitizing affect of the hypoxic radiosensitizer, AK-2123. - Radiotherapy is widely used to combat many cancers. It is known that cells in hypoxic regions of solid tumors are resistant to the effects of ionizing radiation. This is due to a lack of molecular oxygen, which can oxidize and fix free-radical damage in the cellular target material. Because many electron-affinic nitroheterocycles can mimic oxygen, they are effective radiosensitizers of hypoxic cells112). The efficiency of a compound to radiosensitize hypoxic cells depends upon its ability to behave as a one-electron oxidant311). Thus, the radiosensitizing efficiency of an electron-affinic compound towards hypoxic cells, both in vivo and in vitro, depends on its one-electron redox potential311). Although nitrofurans and metronidazole were the first hypoxic cell radiosensitizers1315) to be investigated, the results were disappointing due to the inherent neurological toxicity of the compound. The neurotoxic properties of misonidazole and other nitroimidazoles are directly related to the lipophilic properties of the compounds11). Thus, the dose-limiting side effects of the nitroimidazole have restricted their use in the clinic and lead to a search for compounds16,17) with a similar sensitizing ability and lower toxicity, or with an enhanced sensitizing ability so that it could translate into lower clinical doses with a concomitant reduction of toxic side effects. Hence, many drugs were subsequently designed by judicious tailoring of the substituents on an electron-affinic nitro hetero arene ring. The designed drugs are supposed to be less neurotoxic because of being highly polar, and therefore less able to cross the blood-brain barrier11,12). Among the various nitroheterocyclic hypoxic cell radiosensitizers, sanazole [(N-2-(methoxy ethyl)-2-(3-nitro-1-triazolyl)acetamide], commonly referred to as AK-2123 (scheme 1), has been shown to have selective toxicity for hypoxic cells1,2). It has less toxicity compared to other nitroimidazoles and is currently undergoing phase-III clinical trials. It was found to induce vasoconstriction, leading to hypoxia18). In continuation of our interest in drugs carrying nitro groups19) and considering the importance of AK-2123, we have investigated the transient behavior and various redox reactions occuring in irradiated AK-2123 solutions by a pulse radiolysis technique. In addition, cellular radiosensitization studies by AK-2123 are presented. The one-electron redox potential for the couple AK-2123/AK-2123. at neutral pH was determined. The thus-obtained results were used for a comparison and an analysis in the present study. MATERIALS AND METHODS All of the chemicals and reagents were of AR or GR grade. Spectro-grade 2-propanol and tert-butanol were used. Iolar-grade (purity >99%) gases (N2 or N2O) were used for purging the solutions. Methyl viologen (MV2+) and thymine were obtained from Aldrich. All solutions were prepared in 0.001 mol dm3 phosphate buffer, pH 7 (equimolar mixtures of Na2HPO4.H2O and KH2PO4). Nanopure water (conductivity 0.06 s cm1) was used for making the solutions. The UV-VIS ground-state absorpion spectra were measured with a Hitachi (model 330) spectrophotometer. Bacterial strain and culture conditions Escherichia coli AB1157 cells from a single colony were grown in a Luria broth at 37C overnight. A Luria broth with 1.5% agar was used as a solid medium for plating the bacteria for survival studies. Thiol depletion in bacterial cells After bacterial cells from overnight cultures were harvested, the medium washed off with a sterile buffer (0.067 mol dm3 phosphate) and resuspended in the buffer at a density of (1.0 3.0) 108 cells/ml. Cells were depleted of the intracellular non-protein thiols by treating with NEM at a concentration of 1.0 104 mol dm3 in 0.067 mol dm3 phosphate buffer for one hour at 0C. The possibility of a NEM reaction with several other thiols in proteins cannot be completely neglected. After one hour, cells were centrifuged out and washed with 0.067 mol dm3 phosphate buffer to remove NEM and suspended in the same buffer. Radiolytic procedures The bacterial cell suspensions with and without the NEM treatment were exposed to various doses of 60Co -rays at 0C under a nitrogen atmosphere with a dose rate of 16 Gy min1, as determined by Fricke dosimetry, in the presence and absence of 5.0 103 mol dm3 AK2123 in a 0.067 mol dm3 phosphate buffer. After irradiation the cells were plated on Luria agar plates following serial dilution. The plates were incubated overnight and the colonies counted to determine the survival. It is important to mention here that several of the early compounds that showed hypoxic radiosensitization in bacterial cells never exhibited effectiveness in human tumor cells20). Hence, though the use of bacterial cells for mechanistic studies of sensitization by clinical drugs is not the best choice, it can nevertheless provide some basic information required for carrying out further studies. Pulse radiolysis studies were carried out by irradiating solutions in rectangular quartz cells. The pulse radiolysis set-up consists of an electron linear accelerator (Viritech Ltd. England) capable of giving single pulses of 7 MeV electrons of 50 ns, 500 ns or 2 s duration. Pulses irradiated a sample contained in a 1 cm 1 cm suprasil quartz cuvette held at a distance of approximately 12 cm from the electron beam window, where the beam diameter was approximately 1 cm. The transient changes in the absorbance of the solution caused by the e (...truncated)