Biological Free Radicals

BY MICHAEL J. DAVIES AND GRAHAM S. TIMMINS

1 Introduction and Scope of Review

This review covers recent literature on the use of EPR techniques to investigate the formation and reactions of radicals in biochemical, biological and medical systems during the period 1998 (when this area was last reviewed) to early 2000. It covers both direct EPR spectroscopy and spin trapping studies as well as related techniques; it does not cover metalloprotein systems, DNA damage or spin labelling studies; these topics are covered elsewhere in this volume and the previous volume. Owing to the increasing interest in, and use of, EPR in the biomedical field, this review cannot be all encompassing because of space limitations; we have however endeavoured to cover all the major advances that have occurred during this period, and apologise for any omissions. Emphasis has been placed on novel reactions, and we only cover briefly the large number of studies where spin trapping has been employed in the assessment of putative antioxidant/scavenging compounds, in which the trapping of HO· or O2·- by DMPO* (to give the DMPO-OH or DMPO-OOH adducts) has been employed solely as a competitive 'clock' reaction. The large volume of literature that has developed over the last few years on the trapping of nitric oxide (NO·), which has a wide variety of biological functions, is reviewed briefly, with particular emphasis on EPR methods. The literature covered in this review has been subdivided in a manner similar to that in previous reviews of this area with the sub-sectioning dictated by the exogenous/endogenous compounds or stimuli which result in radical formation, rather than the identity of the radicals so formed.

2 Metal Ions

2.1 Iron. – The origin of the oxygen atom present in HO· generated by the Fenton reaction has been shown by use of 17O labelling and spin trapping with DMPO, to come from the starting H2O2 rather than from water molecules, and that there is no exchange between these species. The intermediacy of ferryl species in tbese reactions could not however be eliminated if the ferryl oxygen atom arose from the labelled H2O2. Other workers have carried out a kinetic analysis of the Fenton reaction by use of DMPO spin trapping, and concluded that the time-dependent loss of DMPO-OH is due to reaction with Fe3+ and not Fe2+. The effect of chelation by oxalate on the rate of reaction of Fe2+ with H2O2 and O2 has been examined by rapid-flow EPR with t-BuOH as a radical scavenger. Using this method, rate constants of 1 × 104 M-1 s-1 and 3.6 M-1 s-1 have been determined for these two processes respectively. Reaction with O2 has been shown to be highly pH dependent consistent with the reactive species being [Fe2+(oxalate)2]2-. Reaction of this complex with H2O2 has also been shown to give rise to CO2-. by use of DMPO as a spin trap. Studies on the significance of reaction of iron complexes with O2versus H2O2 have concluded that when the ratio of O2/H2O2 is >100 then the reaction of Fe2+ with O2 becomes an important route to the formation of detrimental radicals in biological systems. Other studies have examined the effect of a range of agents on the extent of HO· formation by iron-containing systems using spin trapping, primarily with DMPO, as the means of assessing radical yields. Thus α-hydroxyacids were found to significantly enhance HO· formation, whereas other hydroxyacids, ketoacids, organic acids and aldehydes did not; the Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane N,N,N',N'-tetraacetic acid was found to diminish oxidant formation,8 and Lu-Duo-Wei, an extract from Chinese green tea, decreased HO· formation; the effect of this extract was greater than with tea polyphenols alone, suggesting that other compounds also played a role. In contrast, furanone compounds present in soy sauce have been shown to increase DNA damage and enhance HO· formation.

The effect of a series of novel iron chelating agents developed for the potential treatment of β-thalassemia major on HO· formation from the Fenton reaction has been investigated using DMPO; these pyridoxal- and salicylalde-hyde-isonicotinoyl hydrazone compounds decrease radical formation and enhance auto-oxidation of the Fe2+ complex to the Fe3+ form, consistent with these chelators altering the redox potential of the iron couple, but not the radicals formed. The use of spin trapping reactions, with DMPO and PBN as traps for HO· and 1-hydroxyethyl radicals respectively, but with GC/MS as the detection method for the adduct species, has been investigated. It has been reported that after suitable extraction and derivatisation, these adducts can be readily quantified and the technique provides a sensitive assay method for the generation of these species. The use of ethanol with the detection of 1- hydroxyethyl radicals by spin trapping with POBN as a method of examining HO· formation in complex biological systems has continued to be employed. Thus it has been reported that E. coli lacking SOD are not only more susceptible to DNA damage and cell killing by H2O2, but also contain higher levels of intracellular redox-active iron, as judged by enhanced levels of the POBN adducts of 1-hydroxyethyl radicals, generated as a result of HO· formation via Fenton chemistry. Pre-incubation of the bacteria with the iron-chelator desferrioxamine reduced the concentration of spin adducts detected. This technique has also been employed to examine HO· formation in chondrocyte cells stimulated with phorbol esters in the presence of Fe2+ and ethanol; the formation of the observed radicals required the presence of iron, suggesting that a Fenton-type reaction is responsible for radical formation. Similar radical formation was observed with un-stimulated fresh human and rabbit cartilage pieces in the presence of Fe2+; this may be due to metal auto- oxidation. The chelator and dye Quin2 was found to have variable effects depending on the assay system used, with this compound enhancing oxidant formation in a Fe3+/H2O2 system when examined using this POBN/ethanol system as a trap for oxidants.

The mutagenicity of p-aminophenol in E coli has been ascribed to the formation of HO· (as detected by DMPO spin trapping) as a result of interaction of the phenol with Fe3+. The formation of this radical adduct was inhibited by HO· scavengers and catalase suggesting Fenton chemistry as the radical source. In other studies, it has been shown that injection of a number of iron complexes including Fe-EDTA, Fe-desferrioxamine and hemoglobin in to the peritoneum of rats results in superoxide and/or HO· formation as detected by DMPO spin trapping.

The effect of iron-citrate administration on the formation of NO· in mice subjected to exposure to lipopolysaccharide (LPS) has been examined. When the iron complex was administered 6 h after the exposure to LPS an increased concentration of the NO·-iron-diethylthiocarbamate complex was detected 30 min after iron exposure, but no increase in the levels of nitrosylhemoglobin. The authors concluded that this effect arose from an enhanced rate, or extent, of trapping rather than an increase in NO· formation. Later work by this group has shown that when the iron complex is co-administered with the LPS, a decrease in NO· formation is however observed and this has been attributed to an iron-dependent down regulation of the expression of the enzyme iNOS which generates NO·. The level of low-molecular-weight iron present in intact rat hepatocytes has also been probed by use of EPR. In these experiments the hepatocytes were pre-incubated with iron chelators, such as desferrioxamine and deferiprone, and the resulting complex levels quantified by EPR. Ethanol was found to increase low-molecular-weight iron levels, but administration of LPS or gamma-interferon decreased the levels.

2.2 Copper. – The catalysis of damage to biological substrates by Cu(Il) complexes has continued to attract attention. Reaction of DMSO with HO· is known to give rise to methyl radicals and the subsequent trapping of these carbon-centred radicals with POBN has been used as evidence for the formation of HO· in DNA damage induced by mixtures of Cu(II) and biological reducing agents (including ascorbic acid, glutathione and N-Ac-cysteine). A good correlation was observed between the yield of methyl radicals trapped by POBN and the extent of DNA strand scission suggesting that HO· is involved in the genesis of this damage. Related studies have shown that other thiols such as 2-mercaptoethanol can also act as reductants for Cu(II) and hence behave as catalysts for HO· formation from H2O2 (as detected by trapping with DMPO) and DNA damage. The extent of DNA damage was found, not unexpectedly, to depend on the ligands bound to the Cu(II). The role of free HO· in these reactions should, however, be viewed cautiously as it has also been reported that Cu(Il)-peroxide complexes can be formed with H2O2 and that some of these complexes can generate 1O2 which has been detected via oxidation of 2,2,6,6-tetramethyl-4-piperidinol and 2,2,6, 6-tetramethyl-4-piper-idinone to the corresponding nitroxide radicals. The extent of 1O2 formation and the nature of the DNA cleavage patterns were again found to depend on the ligands attached to the Cu(II). In related studies it has been shown that simple mixtures of Cu(II) and sugars (particularly glucosamine, mannosamine and galactosamine) can give rise carbon-centred radicals (trapped with DBNBS) and HO· adduct to DMPO, with these species reported to give rise to DNA strand breaks. These radicals have been ascribed to oxidation of the sugars at neutral pH. The role of Cu and cadmium bound to the metal-chelating protein metallothionein in the catalysis of DMPO/HO· formation has also been explored. With commercial samples of the protein DMPO/HO· formation was detected, but this has been shown to be due to the reaction of Cu ions with H2O2, and that cadmium loaded metallothionein does not give rise to this species; thus cadmium does not appear to catalyse pseudo-Fenton reactions, and the previous positive catalytic effects of metallothioneins are due to Cu bound to this protein.

2.3 Cobalt. – Reaction of Co(II) ions with water have been shown to generate the oxidised species DMPOX from DMPO consistent with the generation of powerful oxidant species. However, this process has been postulated to occur via a Co(I)/H2O2 pseudo-Fenton reaction rather than via a direct reaction of Co(II) with H2O2 with the peroxide arising via superoxide radical formation. The presence of suitable ligands for the Co(II) alters the redox potentials, however, and makes this metal ion into a pseudo-Fenton catalyst with formation of Co(III) ions. Thus the oxidant species generated by Co ions depends dramatically on the ligands to which it is bound via their effect of the redox properties of the metal ion. Similar ligand-dependent Co chemistry has been reported to be involved in metal allergy reactions with both Co and Ni.

2.4 Chromium. – A number of Cr species, particularly Cr(VI) salts, are known human carcinogens and a considerable amount of work has been carried out to elucidate the role of reactive radicals and other species (e.g. Cr(V) complexes) in the actions of these materials. Previous in vivo studies in rats acutely poisoned with Cr(VI) resulted in the detection of a carbon-centred radical adduct to POBN in the bile from these animals. These studies have now been extended to determine the source of this adduct. The coupling constants of the species detected (a(N) 1.571, a(H) 0.29 mT) are very similar to those detected for the pentyl radical adduct derived from metabolism of pentylhydrazine, and from the pentyl radical generated on metabolism of arachidonic acid. The assignment of the observed species to such a radical is supported by the detection of elevated levels of F2 isoprostanes (a marker of lipid oxidation) in the bile from these animals. The effect of Cl-lipoic acid on Cr(VI)-induced damage has been investigated in in vitro studies. This compound, which has been postulated to act as an antioxidant, was observed to reduce Cr(VI) to Cr(V) and result in HO· formation as detected by EPR. Mechanistic studies have shown that the latter radical, which was trapped with DMPO, arises via the formation of superoxide radicals on the basis of oxygen consumption measurements and the effects of catalase and SOD. Thus this supposed antioxidant may act as a exacerbating influence in Cr-induced damage. The effect of Cr ions on gene activation has also been examined in cell cultures, and it has been shown that treatment of cells with Cr(IV) complexes results in activation of the nuclear transcription factor NF-κ B, generation of DNA strand breaks and hydroxylation of 2'-dG; these processes occurred concomitantly with HO· formation as detected by EPR spin trapping using DMPO. Somewhat similar studies with Cr(VI) have concluded that p53-induced apoptosis in human lung cells occurs at relatively long time points after exposure and it was concluded that radical species generated by Cr(VI) are involved in the early stages of apoptosis observed with these cells.

2.7 Other Metal Ions. – EPR studies on the reaction of carcinogenic vanadium salts with mouse epidermal cells have shown that these cells can reduce V(V) to V(IV) in a process that involves NADPH. This process also involves super-oxide radicals as evidenced by DMPO spin trapping; the identity of this adduct was confirmed by use of SOD. These studies provide support for a model where the activation of the transcription factor activator protein (AP-I) is dependent on the generation of superoxide radicals and H2O2 by vanadate, but not HO· formation. The effects of inorganic arsenic ions on gene activation have been studied, and it has been concluded that low concentrations of these ions not only alter gene expression but also generate enhanced levels of oxidant species; the elevated levels of oxidants have been suggested to be responsible for the gene activation. Studies have also been carried out with selenium ions. Though Se is an essential component of the protective enzyme glutathione peroxidase, it has been shown that enhanced levels of such ions can result in a stimulation of the formation of radical formation as detected by a number of methods including EPR spin trapping.

3 Superoxide and Superoxide Dismutases

The role of superoxide dismutases (SOD) in both the stimulation and prevention of radical mediated damage has been the subject of increasing interest particularly with regard to the role of the levels of, and mutations in, this enzyme in various diseases, particularly familial amylotrophic lateral sclerosis. Thus the scavenging action of DMPO has been used to probe the role of Cul Zn SOD in enhancing DNA damage, with the protective effect of DMPO interpreted in terms of the spin trap intercepting HO· generated by Cu released from the damaged enzyme, and it has been shown that transgenic mice with enhanced levels of the Cu/Zn form of the enzyme show increased levels of HO· formation as assessed by DMPO spin trapping. However, more recent studies have questioned the formation of HO· by SOD enzymes. Thus studies using 17O-labelled oxygen and H2O2, have concluded that much of the DMPO-OH observed arises from oxidation of the spin trap and subsequent hydration of the radical-cation rather than direct HO· generation by the enzyme or released metal ions, whereas another study has concluded that the species which oxidises and hydroxylates substrates is a 'bound-HO·' species, and that this enzyme does not release free HO·.