Free radical generation by ultrasound in aqueous solutions of nucleic acid bases and nucleosides: an ESR and spin-trapping study

Direct evidence for the detection of intermediate radicals of nucleic acid constituents induced by ultrasound in argon-saturated aqueous solution is presented. The method of spin trapping with 3,5-dibromo-4-nitrosobenzene sulphonate, which is a water-soluble, non-volatile, aromatic nitroso spin trap, combined with ESR, was used for the detection of sonochemically induced radicals. Spin adducts were also generated by OH radicals produced by UV photolysis of aqueous solution containing H2O2. ESR spectra observed from these photolysis experiments were identical to those after sonolysis. The ESR spectra of the spin adducts suggest that the major spin-trapped radical of thymine and thymidine was the 5-yl radical, and that of cytosine, cytidine, uracil, and uridine was the 6-yl radical. To compare the radicals induced by sonolysis and photolysis, the decay of the ESR spectra of the thymine and thymidine spin adducts was investigated. The decay curves of thymine and thymidine after sonolysis indicated biphasic decay. However, after photolysis the spin adducts from both compounds showed very little decay. These results suggest that the observed spin adducts in the sonolysis of pyrimidine bases and nucleosides were formed by OH radical and H atom addition to the 5,6 double-bond.     

Free Radical Formation by Ultrasound in Aqueous Solutions. A Spin Trapping Study

Our recent spin trapping studies of free radical generation by ultrasound in aqueous solutions are reviewed. The very high temperatures and pressures induced by acoustic cavitation in collapsing gas bubbles in aqueous solutions exposed to ultrasound lead to the thermal dissociation of water vapor into H atoms and OH radicals. Their formation has been confirmed by spin trapping. Sonochemical reactions occur in the gas phase (pyrolysis reactions), in the gas-liquid interfacial region, and in the bulk of the solution (radiation-chemistry reactions). The high temperature gradients in the interfacial regions lead to pyrolysis products from non-volatile solutes present at sufficiently high concentrations. The sonochemically generated radicals from carboxylic acids, amino acids, dipeptides. sugars, pyrimidine bases. nucleosides and nucleo-tides were identified by spin trapping with the non-volatile spin trap 3.5-dibromo-2.6-dideuterio-4-nitrosobenzenesulfonate. At low concentrations of the non-volatile solutes. the spin-trapped radicals produced by sonolysis are due to H atom and OH radical reactions. At higher concentrations of these non-volatile solutes, sonolysis leads to the formation of additional radicals due to pyrolysis processes (typically methyl radicals). A preferred localization of non-volatile surfactants (compared to analogous non-surfactant solutes) was demonstrated by the detection of pyrolysis radicals at 500-fold lower concentrations. Pyrolysis radicals were also found in the sonolysis of aqueous solutions containing only certain nitrone spin traps. The more hydrophobic the spin trap, the lower the concentration at which the pyrolysis radicals can be observed. The effect of varying the temperature of collapsing transient cavities in aqueous solutions of different rare gases and of N₂O on radical yields and on cell lysis of mammalian cells was investigated.     

Effect of ultrasound and ionizing radiation on a sterically hindered cyclic secondary amine: an ESR study.

The possible use of 2,2,6,6-tetramethyl-4-piperidone (TMPone) for the detection of singlet oxygen was investigated by gamma radiolysis and sonolysis of oxygen-saturated aqueous solutions. Formation of 2,2,6,6-tetra-methyl-4-piperidone-N-oxyl (TAN) was observed with both gamma radiolysis and sonolysis with a similar dependence on the concentration of TMPone up to 20 mM and a strong dependence on pH. In oxygen-saturated solutions the sonolysis of TMPone leads to the formation of the cyclic hydroxylamine (approx. 30% of the yield of TAN) while radiolysis does not. In the low pH range (5-6.5) and at high concentrations of OH radical scavengers (azide or formate), TAN is produced by sonolysis but not by radiolysis. Sonolysis of argon-saturated solutions of TMPone produces methyl radicals due to the high-temperature regions of the collapsing cavitation bubbles. The methyl radicals were detected by ESR (electron spin resonance) and spin trapping with 3,5-dibromo-2,6-dideuterio-4-nitroso-benzene sulfonate. Since the reaction of singlet oxygen with TMPone is also strongly dependent on pH, it does not seem likely that TMPone could be used for the detection of singlet oxygen in sonochemistry.     

Sonolysis, radiolysis, and hydrogen peroxide photolysis of pyrimidine derivatives in aqueous solutions: a spin-trapping study

The sonolysis of aqueous solutions of various dihydropyrimidines and substituted pyrimidines was investigated by ESR and spin trapping with the nonvolatile, water soluble spin trap, 3,5-dibromonitrosobenzene sulfonate (DBNBS) and its deuterated analog to examine the possibility of detecting new radicals specifically generated in the high temperature zones produced by collapsing cavitation bubbles. Similar ESR spectra were obtained from sonolysis of argon-saturated aqueous solutions, from uv photolysis of aqueous solutions containing H2O2, and from gamma radiolysis of nitrous oxide saturated solutions, although sonolysis of aqueous solutions leads to the formation of pyrimidine radicals by H atom as well as OH radical addition to the 5,6 double bond of pyrimidines. No evidence for specific new radicals formed in the high temperature regions induced by cavitation could be found. For the reactions of dihydropyrimidines with hydroxyl radicals additional spin adducts could be detected and identified with the spin trap DBNBS compared to 2-methyl-2-nitrosopropane which was used in previous studies; however, for alkylpyrimidines fewer spin adducts were observed. The use of the deuterated analog of DBNBS is helpful for unambiguous radical structure assignment.     

Formation of .OH and .H in aqueous solutions by ultrasound using clinical equipment

Hydroxyl radicals and hydrogen atoms were produced in argon-saturated aqueous solutions exposed to ultrasound using clinical dental equipment. .OH and .H radicals were detected and identified by ESR and were spin trapped with 5,5-dimethyl-1-delta-pyrroline-N-oxide (DMPO) and alpha-4-pyridyl-1-oxide-N-tert-butylnitrone (POBN). The observed ESR spectra were compared with those obtained from sonolysis of argon-saturated water in an ultrasonic bath, from gamma radiolysis of air-saturated water, and from uv photolysis of aqueous hydrogen peroxide solutions.     

A time-resolved electron spin resonance study of the oxidation of ascorbic acid by hydroxyl radical.

Time-resolved electron spin resonance (ESR) spectroscopy for the study of radicals produced by pulse radiolysis is illustrated by a study of the oxidation of ascorbic acid by OH radical in aqueous solution. In basic solution, the direct oxidation product, the ascorbate mono-anion radical, is formed within less than 2 mus of the radiolysis pulse. In acid solutions (pH 3(-4.5), N(2)O:saturated) three radicals are initially formed, the ascorbate mono-anion radical, an OH adduct seen also in steady-state ESR experiments, and an OH adduct at C2 with the main spin density at C3 of the ring. The first OH adduct decays with an initial half-life of about 100 mus, probably by biomolecular reaction. The second OH adduct, which shows one hyperfine splitting about a(H) = 24.4 +/- 0.3 G and g = 2.0031 +/- 0.0002, decays with a half-life of about 10 mus. On this same time scale the concentration of the ascorbate radical approximately doubles. It is concluded that the adduct at C2, but not the other adduct, loses water rapidly to form the ascorbate radical.     

Spin trapping of precursors of thymine damage in X-irradiated DNA

A spin-trapping method combined with ESR spectroscopy was utilized to obtain evidence for the presence of precursor radicals leading to damage in X-irradiated DNA. Two technical improvements were introduced to the conventional spin-trapping method to make possible its application to large molecules such as DNA: prior to X irradiation, sonolysis of aqueous DNA solution by 19.5-kHz ultrasound was made to get a highly concentrated DNA solution and to lower the viscosity of the solution; after precursor radicals in X-irradiated DNA were trapped by a spin-trapping reagent, the DNA was digested to oligonucleotides by DNase I to get an ESR spectrum with a well-resolved hyperfine structure. Thus, it was recognized that the ESR spectrum obtained after X irradiation of the aqueous solution containing DNA and the nitroso spin-trapping reagent 2-methyl-2-nitrosopropane consisted of at least three sets of signals in the DNA. Identification of free radicals was made by comparing the spectrum with that of thymidine, which was precisely examined by a spin-trapping method combining two kinds of spin traps (nitroso and nitrone compounds) with liquid chromatography. As a result, all the signals were identified as the spin adducts of radicals produced at the thymine base moiety of DNA. The 5-hydroxy-5,6-dihydrothymin-6-yl radical was identified as a precursor of 5,6-dihydroxy-5,6-dihydrothymine (thymine glycol), the 6-hydroxy-5,6-dihydrothymin-5-yl radical as a precursor of 6-hydroxy-5,6-dihydrothymine, and the 5-methyleneuracil radical as a precursor of 5-(hydroxymethyl) uracil.     

Scavenging of OH radicals produced in the sonolysis of water

The yield of hydrogen peroxide in the sonication of argon-saturated water was studied in the presence of various solutes. The efficiency of OH radical scavenging is expressed by the reciprocal value of C 1/2, the solute concentration at which the H2O2 yield is decreased by 50 per cent. C 1/2 ranges over several orders of magnitude. It is not related to the specific reactivity towards OH in homogeneous solution. However, it is correlated to the hydrophobicity of the solutes. The competition of I- and a second solute for OH was also studied. The competition between I- and HCO2- follows similar kinetics as in homogeneous solution. However, many other solutes compete in the manner which would be expected if radical scavenging occurred in different phases. The effects are explained in terms of OH radical formation in gaseous argon bubbles, combination of OH radicals to form H2O2 in an interfacial area, and enrichment of hydrophobic solutes in the bubbles.     

Mammalian cells are not killed by DNA single-strand breaks caused by hydroxyl radicals from hydrogen peroxide

Cell killing by ionizing radiation has been shown to be caused by hydroxyl free radicals formed by water radiolysis. We have previously suggested that the killing is not caused by individual OH free radicals but by the interaction of volumes of high radical density with DNA to cause locally multiply damaged sites (LMDS) (J. F. Ward, Radiat. Res. 86, 185-195, 1985). Here we test this hypothesis using hydrogen peroxide as an alternate source of OH radicals. The route to OH production from H2O2 is expected to cause singly damaged sites rather than LMDS. Chinese hamster V79-171 cells were treated with H2O2 at varying concentrations for varying times at 0 degree C. DNA damage produced intracellularly was measured by alkaline elution and quantitated in terms of Gray-equivalent damage by comparing the rate of its elution with that of DNA from gamma-irradiated cells. The yield of DNA damage produced increases with increasing concentration of H2O2 and with time of exposure. H2O2 is efficient in producing single-strand breaks; treatment with 50 microM for 30 min produces damage equivalent to that formed by 10 Gy of gamma irradiation. In the presence of a hydroxyl radical scavenger, dimethyl sulfoxide (DMSO), the yield of damage decreases with increasing DMSO concentration consistent with the scavenging of hydroxyl radicals traveling an average of 15 A prior to reacting with the DNA. In contrast to DNA damage production, cell killing by H2O2 treatment at 0 degree C is inefficient. Concentrations of 5 X 10(-2) M H2O2 for 10 min are required to produce significant cell killing; the DNA damage yield from this treatment can be calculated to be equivalent to 6000 Gy of gamma irradiation. The conclusion drawn is that individual DNA damage sites are ineffectual in killing cells. Mechanisms are suggested for killing at 0 degree C at high concentrations and for the efficient cell killing by H2O2 at 37 degrees C at much lower concentrations.     

The role of the DMPO-hydrated electron spin adduct in DMPO-*OH spin trapping

Time-resolved in situ radiolysis ESR (electron spin resonance, equivalently EPR, electron paramagnetic resonance) studies have shown that the scavenging of radiolytically produced hydroxyl radical in nitrous oxide-saturated aqueous solutions containing 2 mM DMPO is essentially quantitative (94% of the theoretical yield) at 100 micros after the electron pulse [1]. This result appeared to conflict with earlier results using continuous cobalt-60 gamma radiolysis and hydrogen peroxide photolysis, where factors of 35 and 33% were obtained, respectively [2,3]. To investigate this discrepancy, nitrogen-saturated aqueous solutions containing 15 mM DMPO were cobalt-60 gamma irradiated (dose rate = 223 Gy/min) for periods of 0.25-6 min, and ESR absorption spectra were observed approximately 30 s after irradiation. A rapid, pseudo-first-order termination reaction of the protonated DMPO-hydrated electron adduct (DMPO-H) with DMPO-OH was observed for the first time. The rate constant for the reaction of DMPO-H with DMPO-OH is 2.44 x 10(2) (+/- 2.2 x 10(1)) M(-1) s(-1). In low-dose radiolysis experiments, this reaction lowers the observed yield of DMPO-OH to 44% of the radiation-chemical OH radical yield (G = 2.8), in good agreement with the earlier results [2,3]. In the absence of the DMPO-H radical, the DMPO-OH exhibits second-order radical termination kinetics, 2k(T) = 22 (+/- 2) M(-1) s(-1) at initial DMPO-OH concentrations > or = 13 microM, with first-order termination kinetics observed at lower concentrations, in agreement with earlier literature reports [4].     

EPR Spin Trapping Study of the Decomposition of Azo Compounds in Aqueous Solutions by Ultrasound: Potential for Use as Sonodynamic Sensitizers for Cell Killing

Sonodynamic therapy, a promising new approach to cancer treatment, is based on synergistic cell killing by combination of certain drugs (sonosensitizers) and ultrasound. Although the mechanism of sonodynamic action is not understood, the role of free radicals produced from sonosensitizers by ultrasound is implicated. In this work, we studied formation of free radicals during the decomposition of several water-soluble azo compounds by 50 kHz ultrasound in aqueous solutions. Using the spin trap 3, 5-dibromo-4-nitrosobenzene sulfonate (DBNBS) tertiary carbon-centered radicals from 2, 2'-azobis (N,N'-dimethyl-eneisobutyramidine) dihydrochloride (VA-044), 2-(carbamoylazo)-isobutyronitrile (V-30), and 2, 2'-azobis (2-amidinopropane) dihydrochloride (AAPH) and CH3 radicals from 1, 1'-azobis (N,N'-dimethylformamide) (ADMF) were detected in argonsaturated solutions and the corresponding oxygen-centered radicals (alkoxyl and peroxyl) from VA-044, V-30, and AAPH were identified using the spin trap 5, 5'-dimethyl-l-pyrroline-N-oxide (DMPO) in aerated sonicated solutions. No free radicals from 4, 4'-dihydroxyazobenzene-3, 3'-dicarboxylic acid, disodium salt (DHAB) could be found in either system. While VA-044 and AAPH could also be readily decomposed by heat (42.5°C and 80°C), V-30 decomposition only occurred in the ultrasound-exposed solutions. The most likely mechanism of decomposition of azo compounds by ultrasound is their thermolysis in the heated shell of the liquid surrounding ca vita ting bubbles driven by ultrasound and/or by pyrolysis inside these bubbles. Experiments using scavengers of ·OH and ·H, which are produced by sonolysis in aqueous solutions, demonstrated that these radicals are not involved in the ultrasound-mediated radical production from the azo compounds. Due to the known cytotoxic potential of free radicals produced from azo compounds, the use of these compounds as ultrasound sensitizers appears to be a promising approach for sonodynamic cell killing.     

E.S.R. studies of free radicals derived from reaction of OH with uracil and thymine

Short-lived free radicals produced in N2O-saturated aqueous solutions of uracil and thymine have been studied using the in situ radiolysis steady-state e.s.r. method. Radical formed in alkaline aqueous solutions by OH addition to either positions C(5) or (6) were observed. Mechanisms for the formation of transient species were derived. The spin density distribution of the unpaired electron was calculated by means of the INDO method.     

Use of EPR and FTIR to detect biological effects of ultrasound and microbubbles on a fibroblast cell line

Structural and functional effects of exposing murine fibroblasts (NIH 3T3) to therapeutic ultrasound at 1 MHz frequency are described. These bioeffects can be attributed to the formation of free radical species by sonolysis of water. When cavitation occurs, dissociation of water vapor into H atoms and OH radicals is observed; these H atoms and OH radicals combine to form H₂, H₂O₂, and HO₂. The radicals can chemically modify biomolecules, for example enzymes, DNA, and lipids. Generation of free radicals during exposure to ultrasound with or without encapsulated microbubbles (contrast agents) was studied by use of electron paramagnetic resonance with DMPO spin trapping. Recently the potential for possible use of these microbubbles in gene therapy has been investigated, because of the ability of the stabilized microbubbles to release their content when exposed to ultrasound. Structural changes were studied by Fourier-transform infrared spectroscopy, and induction of possible genotoxic damage by exposure of the cells to therapeutic ultrasound at 1 MHz frequency with our experimental device was verified by use of the cytokinesis-block micronucleus assay.     

Hot spot kinetics of the sonolysis of aqueous acetate solutions

Water and acetate solutions were irradiated under argon by 300 kHz ultrasonic waves. Oxygen was found to be generated besides the products H2 and H2O2, already known. In the presence of acetate the O2 yield decreased rapidly while that of H2O2 decreased more slowly. Succinic acid was found as a product of the attack of OH radicals on acetate. Appreciable amounts of glyoxylic and glycolic acid and smaller amounts of formaldehyde and carbon dioxide were also detected. They resulted from the reaction of sonolytically generated oxygen with CH2CO2- radicals, produced upon attack of OH on acetate. Methane was a minor product of sonolysis. At acetate concentrations above 0.4 mol dm-3 CO2 and CO became the predominant products of sonolysis. This is explained by a second kind of action of ultrasound on dissolved acetate, i.e. by a thermal decomposition. This decomposition is possibly facilitated by radical attack on acetate. The results are discussed in terms of a 'structured hot spot' model, in which three regions for the occurrence of chemical reactions are postulated: a hot gaseous nucleus, an interfacial region with radial gradient in temperature and local radical density; and the bulk solution at ambient temperature.     

EPR characterization of free radical intermediates formed during ultrasound exposure of cell culture media

Free radicals and/or hydrogen peroxide produced by exposure of cells to ultrasound are potentially cytotoxic and mutagenic. The formation and type of free radical species can be substantially modulated by the chemical composition of the media in which the ultrasound exposures of cells are carried out. In the current study, we examined the free radical intermediates formed during ultrasound exposure of a typical cell culture medium (RPMI-1640); the dominant free radicals that were identified by spin trapping were derived from the hydrophobic amino acids Trp, Leu, and Phe, and were formed by hydrogen abstraction from these amino acids. Compared to exposures in phosphate-buffered saline, the yield of *OH radicals and H2O2 was significantly reduced in the cell culture medium, glucose (the main organic component in the medium), and the hydrophobic amino acids (Trp, Phe, Tyr, Leu, Val, Met) being chiefly responsible for this effect. In contrast, other nonhydrophobic amino acids did not contribute significantly to the *OH or H2O2 decrease. These findings are consistent with the accumulation of hydrophobic solutes at the liquid-gas interface of the collapsing cavitation bubbles resulting in increased efficiency of radical scavenging.     

DMPO-Alkyl radical spin trapping: an in situ radiolysis steady-state ESR study

Short-lived free radicals formed in the reaction of 11 substrates and radiolytically produced hydroxyl radicals were trapped successfully with 5, 5-dimethyl-1-pyrroline-N-oxide (DMPO) in dilute aqueous solution. The in situ radiolysis steady-state ESR spectra of the spin adducts were analyzed to determine accurate ESR parameters for these spin adducts in a uniform environment. Parent alkyl radicals include methyl, ethyl, 1-propyl and 2-propyl (1-methylethyl). Hydroxyalkyl parent radicals were hydroxymethyl, hydroxyethyl, 2-hydroxy-2-propyl (1-methyl-1-hydroxyethyl), 1-hydroxypropyl and 2-hydroxy-2-methylpropyl. Carboxyl radical (carbon dioxide anion, formate radical) and sulfite anion radical were the sigma radicals studied. The DMPO spin adduct of 1-propyl was identified for the first time. For most spin adducts, g factors were also determined for the first time. In DMPO spin adducts of hydroxyalkyl radicals, nitrogen and C(2)-proton hyperfine coupling constants are smaller than those of alkyl radical adducts; the hydroxyalkyl spin adducts possess larger g values than their unsubstituted counterparts. These changes are ascribed to the spread of pi conjugation to include the hydroxyl group. Strong evidence of spin addend-aminoxyl group interaction can be seen in the asymmetrical line shapes in the hydroxyethyl and the hydroxypropyl spin adducts.     

The hydroxyl free radical reactions of ascorbyl palmitate as measured in various in vitro models.

The OH(*) free radical scavenging properties of ascorbyl palmitate (AP), water-solubilized in the presence of a surfactant (Brij 35), were tested in various systems: (1) The inhibition of polymerization of bovine serum albumin by OH(*) free radicals generated by the Fenton reaction indicated AP exerts a considerable protective effect against polymerization by scavenging the OH(*) free radicals. (2) ESR spin trapping comparisons of DMPO with AP were conducted. Using the Fenton reaction as a source of OH(*) free radicals, AP was 1 order of magnitude faster in scavenging these radicals than DMPO. (3) Oxidative modification of BSA by (60)Co-gamma irradiation of 80 krad, results in a strong increase in protein carbonyl content. AP inhibits carbonyl formation very efficiently, indicating that AP may be utilized as a biological OH(*) free radical scavenger in human therapy.     

OH radicals and oxidizing products in the gamma radiolysis of water

The production of OH radicals in the gamma radiolysis of water has been examined with radical scavenger techniques employing formic acid. OH radical yields were found to vary from 2.4 radicals/100 eV at the low scavenger concentration limit to 4.2 at a formic acid concentration of 3 M. An inverse Laplace transform technique was applied to the scavenger concentration dependence to obtain the temporal dependence of OH radicals in pure water. It was found that the relative decrease in OH radical yields from 200 ps to 3 ns was virtually the same for the transform of the scavenger data and the directly measured time-resolved results. The absolute yields for the time-resolved experiments are about 10% higher than expected from the present results with scavengers. The agreement can be considered to be good, and reasons for the observed difference are given. Approximately 40% of the OH radicals produced lead to the formation of hydrogen peroxide, which is the only other major oxidizing species in the gamma radiolysis of water. The net water decomposition for gamma rays was found to vary from an initial value of 5.6 +/- 0.3 molecules/100 eV to 3.8 +/- 0.2 molecules/100 eV at 1 micros.     

E.s.r. of spin-trapped radicals in gamma-irradiated aqueous solutions of nucleic acids and their constituents.

The gamma-radiolysis of de-aerated neutral aqueous solutions of uracil, thymine, cytosine and of the corresponding nucleosides and nucleotides and of calf-thymus DNA was investigated. For uracil and thymine, the U.V. photolysis of aqueous solutions containing H2O2 was also studied. The short-lived radicals were spin-trapped by tert-nitrosobutane and identified by electron-spin-resonance spectroscopy. For all compounds two or more radicals were observed, and these could be distinguished by following the thermal decay of the spin adducts. Radicals formed by the addition of H or OH at the C(5) or C(6) positions of the pyrimidine derivatives were observed in all cases. Sodium formate was used as a scavenger for H and OH to identify the radicals formed by eaq-. Spin-trapped radicals in gamma-irradiated aqueous solutions of polynucleotides exhibited broad e.s.r. lines. For DNA gel, additional narrow lines due to scission products were also found.     

Possible role for metallothionein in protection against radiation-induced oxidative stress. Kinetics and mechanism of its reaction with superoxide and hydroxyl radicals

Rabbit liver metallothionein-1 (Mr 6500), which contains zinc and/or cadmium ions, appears to scavenge free hydroxyl (.OH) and superoxide (O-.2) radicals produced by the xanthine/xanthine oxidase reaction much more effectively than bovine serum albumin (Mr 65 000) which was used as a control. Kinetic competition studies between metallothionein and either a spin trap for .OH or ferricytochrome c for O-.2 radicals, gave bimolecular rate constants of the order of kOH/MT approximately equal to 10(12) M-1 X s-1 and kO-2/MT approximately equal to 5 X 10(5) M-1 X s-1, respectively. The former value suggests that all 20 cysteine sulfur atoms are involved in this quenching process and that they all act in the diffusion control limit. The aerobic radiolysis of an aqueous solution of metallothionein, generating O-.2 and .OH radicals, induced metal ion loss and thiolate oxidation. These effects could be reversed by incubation of the irradiated protein with reduced glutathione and the appropriate bivalent metal ion. Metallothionein appears to be an extraordinarily efficient .OH radical scavenger even when compared to proteins 10-50-times its molecular weight. Moreover, hydroxyl radical damage to metallothionein appears to occur at the metal-thiolate clusters, which may be repaired in the cell by reduced glutathione. Metallothionein has the characteristics of a sacrificial but renewable cellular target for .OH-mediated cellular damage.