Allan J. Davison

Protection by β-carotene and related compounds against oxygen-mediated cytotoxicity and genotoxicity: Implications for carcinogenesis and anticarcinogenesis

beta-Carotene protects against photooxidative dermatitis in porphyric humans and mice by quenching of photoactivated species. Other actions of beta-carotene in vivo are explained on the basis of its ability to scavenge free radicals in vitro. For example, in guinea pigs treated with CCl4, beta-carotene decreases pentane and ethane production. Epidemiological studies link low serum beta-carotene levels to elevated risk of lung and other cancers, and in intervention trials, beta-carotene diminishes preneoplastic lesions. However, the dose/response relationships are not well established, and antineoplastic mechanisms await clarification. Given a radical quenching mechanism, beta-carotene should block tumor promotion, but more typically the site of action is progression and an even later role in invasion has not been ruled out. Some antineoplastic actions of carotenoids (such as increased rejection of fibrosarcomas in mice) are attributed to immunoenhancement; others may reflect conversion to retinoids and subsequent gene regulation. Carotenoids other than beta-carotene may act at an earlier stage of carcinogenesis or be more effective as anticarcinogens at certain target sites. As scavengers of hydroxyl radicals, canthaxanthin and astaxanthin are more effective than beta-carotene. Canthaxanthin is sometimes more effective than beta-carotene in chemoprevention, but it is sometimes completely ineffective. Lycopene quenches singlet oxygen more than twice as effectively as beta-carotene. However, the antineoplastic actions of lycopene or astaxanthin remain untested. Explorations of the interactions of carotenoids with other nutrients are just beginning. Dietary fat increases absorption of carotene but decreases antineoplastic effectiveness. Research is hampered by technical problems, including the unavailability of rigorous controls, the instability of carotenoids, and the heterogeneous phase structure induced by hydrophobic compounds in aqueous media. Areas of current controversy and promising approaches for future research are identified.

Intermediates in the aerobic autoxidation of 6-hydroxydopamine: Relative importance under different reaction conditions

Autoxidation of 6-hydroxydopamine (6-OHDA) proceeds through a balanced network of: transition metal ions, superoxide, hydrogen peroxide, hydroxyl radicals, and other species. The contribution of each to the reaction mechanism varies dramatically depending upon which scavengers are present. The contribution of each propagating intermediate increases when the involvement of others is diminished. Thus, superoxide (which is relatively unimportant when metal ions can participate) dominates the reaction when transition metal ions are bound (especially at higher pH), and it becomes essential in the simultaneous presence of catalase plus chelators. Transition metal ions participate more if superoxide is excluded; hydrogen peroxide becomes more important if both .O2- and metal ions are excluded; and hydroxyl radicals contribute more to the reaction mechanism if both H2O2 and .O2- are excluded. Superoxide dismutase inhibited strongly, by two distinct mechanisms: a high affinity mechanism (less than 13% inhibition) at catalytically effective concentrations, and a low affinity mechanism (almost complete inhibition at the highest concentrations) which depends upon both metal binding and catalytic actions. In the presence of DETAPAC catalytic concentrations of superoxide dismutase inhibited by over 98%. Conversely, metal chelating agents inhibited strongly in the presence of superoxide dismutase. When present alone they stimulated (like EDTA), inhibited (like desferrioxamine), or had little effect (like DETAPAC). Catalase which stimulated slightly but consistently (less than 5%) when added alone, inhibited 100% in the presence of superoxide dismutase + DETAPAC. However, in the absence of DETAPAC, catalase decreased inhibition by superoxide dismutase, yielding a 100% increase in reaction rate. Hydroxyl scavengers (formate, mannitol or glucose) alone produced little or no (less than 10%) inhibition, but inhibited by 30% in the presence of catalase + superoxide dismutase. Paradoxically, they stimulated the reaction in the presence of catalase + superoxide dismutase + DETAPAC.

DNA-breaking versus DNA-protecting activity of four phenolic compounds in vitro

Given the paradoxical effects of phenolics in oxidative stress, we evaluated the relative pro-oxidant and antioxidant properties of four natural phenolic compounds in DNA nicking. The phenolic compounds differed dramatically in their ability to nick purified supercoiled DNA, with the relative DNA nicking activity in the order: 1,2,4-benzenetriol (100% nicking) > gallic acid > caffeic acid > gossypol (20% nicking). Desferrioxamine (0.02 mM) decreased DNA strand breakage by each phenolic, most markedly with gallate (85% protection) and least with caffeic acid (26% protection). Addition of metals accelerated DNA nicking, with copper more effective (∼5-fold increase in damage) than iron with all four phenolics. Scavengers revealed the participation of specific oxygen-derived active species in DNA breakage. Hydrogen peroxide participated in all cases (23–90%). Hydroxyl radicals were involved (32–85%), except with 1,2,4-benzenetriol. Superoxide participated (81–86%) with gallic acid and gossypol, but not with caffeic acid or 1,2,4-benzenetriol. With 1,2,4-benzenetriol, scavengers failed to protect significantly except in combination. Thus, in the presence of desferrioxamine, catalase or superoxide dismutase inhibited almost completely. When DNA breakage was induced by Fentons reagent (ascorbate plus iron) the two catechols (caffeic acid and gossypol) were protective, whereas the two triols (1,2,4-benzenetriol and gallic acid) exacerbated damage.

Hypothesis: etiology of atherosclerosis and osteoporosis: are imbalances in the calciferol endocrine system implicated?

Atherosclerosis and osteoporosis are currently considered unrelated diseases. Osteoporosis involves bone calcium (Ca) loss and predominantly affects females after menopause. Atherosclerosis is an illness predominantly affecting males, and is primarily characterized by abnormal lipid metabolism. However, pathological calcification of the arterial wall is an underlying feature of atherosclerosis. Ca homeostasis is thus important in atherosclerosis as well as in osteoporosis. Men also develop osteoporosis although at a later age than women, and, as osteoporosis progresses in women, there is an accompanying calcification of arteries leading to increased incidence of atherosclerosis in aging women. Thus, during old age, both atherosclerosis and osteoporosis are prevalent in both males and females. The dramatic increase in atherosclerosis among women as they develop osteoporosis suggests that the two illnesses may be more closely related than previously realized. The use of vitamin D as a food supplement coincides with epidemic onsets of atherosclerosis and osteoporosis, and excess vitamin D induces both conditions in humans and laboratory animals. These observations suggest a role for chronic vitamin D excess in the etiology of the two illnesses. Magnesium (Mg) deficiency, nicotine, and high dietary cholesterol are contributing factors that accentuate adverse effects of vitamin D.

Effects of metals, ligands and antioxidants on the reaction of oxygen with 1,2,4-benzenetriol.

1,2,4-Benzenetriol is an active metabolite of the human leukemogen benzene that reacts rapidly with molecular oxygen (O2). The mechanism of autoxidation of benzenetriol is scantily characterized, and little is known of the effects of metals, metal chelators, radical scavengers, and antioxidants on the rate of reduction of O2. Here, we report that catalytic amounts of Cu2+ and Fe3+ accelerated the oxidation of benzenetriol (250 mu M) in a dose-dependent manner. Fe3+ (50 mu M) increased the rate of autoxidation by 91%, and Cu2+ (10 mu M) increased it 11-fold. In the absence of added metals, superoxide dismutase inhibited and desferrioxamine stimulated the autoxidation. In the Cu2+ -catalyzed reaction, superoxide dismutase neither inhibited nor stimulated, while desferrioxamine abolished the catalysis by Cu2+. In the presence of Fe3+, superoxide dismutase slowed the reaction, but desferrioxamine, surprisingly, did not. The presence of both superoxide dismutase and desferrioxamine blocked the autoxidation, either in the presence or absence of metals. We conclude: (1) superoxide is a propagator of sequential one-electron transfer reactions in the absence of added metals; (2) addition of Cu2+, unlike Fe3+, removes the dependence of the reaction on propagation by superoxide, presumably changing the radical-propagated chain reaction to a concerted two-electron transfer; (3) the further addition of desferrioxamine restores superoxide-dependent propagation. Taken with our previous data on the genotoxicity of benzenetriol, these findings have implications regarding a role for transition metals in the carcinogenicity of benzene.

Multiple actions of superoxide dismutase: Why can it both inhibit and stimulate reduction of oxygen by hydroquinonez?

Superoxide dismutase can either inhibit or stimulate autoxidation of different hydroquinones, suggesting multiple roles for O2.-. Inhibitory actions of superoxide dismutase include termination of O2.(-)-propagated reaction chains and metal chelation by the apoprotein. Together, chelation of metals and termination of O2.(-)-propagated chains can effectively prevent reduction of oxygen. Chain termination by superoxide dismutase can thus account for negligible accumulation of H2O2 without invoking a superoxide:semiquinone oxidoreductase activity for this enzyme. One stimulatory action of superoxide dismutase is to decrease thermodynamic limitations to reduction of oxygen. Whether superoxide dismutase inhibits or accelerates an autoxidation depends on the reduction potentials of the quinone and the availability of metal coordination for inner sphere electron transfers.

The relative effectiveness of .OH, H2O2,O2−, and reducing free radicals in causing damage to biomembranes: A study of radiation damage to erythrocyte ghosts using selective free radical scavengers

The relative effectiveness of oxidizing (.OH, H2O2), ambivalent (O2-) and reducing free radicals (e- and CO2-) in causing damage to membranes and membrane=bound glyceraldehyde-3-phosphate dehydrogenase of resealed erythrocyte ghosts has been determined. The rates of damage to membrane-bound glyceraldehyde-3-phosphate dehydrogenase (R(enz)) were measured and the rates of damage to membranes (R(mb)) were assessed by measuring changes in permeability of the resealed ghosts to the relatively low molecular weight substrates of glyceraldehyde-3-phosphate dehydrogenase. Each radical was selectively isolated from the mixture produced during gamma-irradiation, using appropriate mixtures of scavengers such as catalase, superoxide dismutase and formate. .OH, O2- and H2O2 were approximately equally effective in inactivating membrane-bound glyceraldehyde-3-phosphate dehydrogenase, while e- and CO2- were the least effective. R(enz) values of O2- and H2O2 were 10-times and of .OH 15-times that of e-. R(mb) values were quite similar for e- and H2O2 (about twice that of O2-), while that of .OH was 3-times that of O2-. Hence, with respect to R(mb): .OH greater than e- = H2O2 greater than O2-, and with respect to R(enz): .OH greater than O2- = H2O2 much greater than e-. The difference between the effectiveness of the most damaging and the least damaging free radicals was more than 10-fold greater in damage to the enzyme than to the membranes. Comparison between H2O2 added as a chemical reagent and H2O2 formed by irradiation showed that membranes and membrane-bound glyceraldehyde-3-phosphate dehydrogenase were relatively inert to reagent H2O2 but markedly susceptible to the latter.

6-Hydroxydopamine does not reduce molecular oxygen directly, but requires a coreductant☆

The autoxidation of 6-hydroxydopamine (6HODA) was virtually blocked (k2 less than 10(-15) M-1 S-1 at pH 8.0, ionic strength 0.04) by the simultaneous presence of diethylenetriaminepentaacetic acid (DTPA), catalase, and superoxide dismutase (SOD). No quinone product or oxygen consumption was detectable after 12 min under these conditions. Thus, if 6HODA is to react with molecular oxygen at a measurable rate, some other redox species is required as a coreductant. The subsequent addition of formate or mannitol proved capable of overcoming the total inhibition induced by the mixture of catalase, SOD, and DTPA. The simplest interpretation of the data is that most of the autoxidation of 6HODA, as commonly observed, involves successive reduction of a series of metal-bound species of oxygen; the actual transfer of electrons occurring within a ternary reductant-metal-oxygen transition state.

Interactions between metals, ligands, and oxygen in the autoxidation of 6-hydroxydopamine: mechanisms by which metal chelation enhances inhibition by superoxide dismutase

Transition metal ions and superoxide participate in different autoxidations to a variable extent. In the reaction of 6-hydroxydopamine (6-OHDA) with oxygen at pH 7.0 or 8.0, addition of 5 to 300 U/ml superoxide dismutase inhibited autoxidation by up to 96% at the highest concentrations. Superoxide dismutase at concentrations of 5-20 U/ml inhibited by less than 40% when present alone, but inhibited by over 99% in the presence of desferrioxamine or histidine. EDTA also enhanced the inhibition by 20 U/ml superoxide dismutase to 86%, even though EDTA accelerated the autoxidation of 6-OHDA when present alone or with desferrioxamine. In contrast, other ligands, such as ADP or phytic acid, had little or no effect on inhibition by superoxide dismutase. Proteins such as albumin, cytochrome oxidase, or denatured superoxide dismutase also enhanced inhibition by active superoxide dismutase from less than 40% to over 90%. Evidently, in the presence of redox active metals, autoxidation occurs by inner sphere electron transfer, presumably within a ternary 6-OHDA.metal.oxygen complex. This mechanism does not involve free O2-. and is not inhibited by superoxide dismutase. On the other hand, the presence of certain ligands (including proteins) diminishes the ability of trace metals to exchange electrons with 6-OHDA or oxygen by an inner sphere mechanism. These ligands render autoxidation dependent on propagation by O2-. and therefore inhibitable by superoxide dismutase. Previously conflicting reports that superoxide dismutase alone inhibits 6-OHDA autoxidation are thus explicable on the basis that at sufficient concentration the apoprotein coordinates trace metals in such a way to preclude inner sphere metal catalysis.

Vanadate-induced gene expression in mouse C127 cells: roles of oxygen derived active species.

An underinvestigated aspect of the mitogenic and cell regulatory actions of vanadium is the regulation of gene expression. Among the fifteen cellular genes studied in cultured mouse C127 cells, vanadium (as 10 μM sodium vanadate) increased levels of mRNA of the actin and c-Ha-ras to four times control values. These increases represented de novo synthesis of mRNA, since they were inhibited by actinomycin D. Vanadate did not increase mRNA corresponding to c-src, c-mos, c-myc, p53, HSP70, pODC or RB genes, and expression of c-erb A, c-erb B, c-sis and c-fes genes was undetectable whether vanadium was present or not. Expression of a third gene affected by vanadium, c-jun, was augmented by addition of a reductant or oxidant together with the vanadate. Addition of NADH (marginally effective on its own) or H2O2 (effective alone) dramatically enhanced the effect of vanadate on c-jun gene expression. Catalase inhibited the effect of NADH partly. The vanadate-stimulated expression of actin and c-Ha-ras mRNA were unaffected by oxidants, reductants, metal chelators, or anti-oxidant enzymes. Evidently vanadate acts by two separate mechanisms on these two categories of genes. The alternate hypothesis that the actions of vanadate on actin and c-Ha-ras were mediated by a protein kinase cascade was inconsistent with the following observations. Neither insulin nor epidermal growth factor increased mRNA levels of c-Ha-ras or actin gene. Neither genistein (a tyrosine kinase inhibitor) nor pretreatment with 12-O-tetradecanoylphorbol-13-acetate blocked the actions of vanadate on these genes. Clearly the biological actions of vanadium depend in part on altered expression of genes. Since two of the genes are proto-oncogenes, this mechanism is potentially relevant to the mitogenic responses of cells to vanadium.