Robert A. Floyd

Role of oxygen free radicals in carcinogenesis and brain ischemia.

Even though oxygen is necessary for aerobic life, it can also participate in potentially toxic reactions involving oxygen free radicals and transition metals such as Fe that damage membranes, proteins, and nucleic acids. Oxygen free radical reactions and oxidative damage are in most cases held in check by antioxidant defense mechanisms, but where an excessive amount of oxygen free radicals are produced or defense mechanisms are impaired, oxidative damage may occur and this appears to be important in contributing to several pathological conditions including aging, carcinogenesis, and stroke. Several newer methods, such as in vivo spin‐trapping, have become available to monitor oxygen free radical flux and quantitate oxidative damage. Using a combination of these newer methods collectively focused on one model, recent results show that oxidative damage plays a key role in brain injury that occurs in stroke. Subtle changes, such as oxidative damage‐induced loss of glutamine synthetase activity, may be a key event in stroke‐induced brain injury. Oxygen free radicals may play a key role in carcinogenesis by mediating formation of base adducts, such as 8‐hydroxyguanine, which can now be quantitated to very low levels. Evidence is presented that a new class of free radical blocking agents, nitrone spin‐traps, may help not only to clarify if free radical events are involved, but may help prevent the development of injury in certain pathological conditions.—Floyd, R. A. Role of oxygen free radicals in carcinogenesis and brain ischemia. FASEB J. 4: 2587‐2597, 1990.

Reactive oxygen species, cell signaling, and cell injury

Oxidative stress has traditionally been viewed as a stochastic process of cell damage resulting from aerobic metabolism, and antioxidants have been viewed simply as free radical scavengers. Only recently has it been recognized that reactive oxygen species (ROS) are widely used as second messengers to propagate proinflammatory or growth-stimulatory signals. With this knowledge has come the corollary realization that oxidative stress and chronic inflammation are related, perhaps inseparable phenomena. New pharmacological strategies aimed at supplementing antioxidant defense systems while antagonizing redox-sensitive signal transduction may allow improved clinical management of chronic inflammatory or degenerative conditions, including Alzheimers disease. Introduction of antioxidant therapies into mainstream medicine is possible and promising, but will require significant advances in basic cell biology, pharmacology, and clinical bioanalysis.

Oxidative stress in brain aging. Implications for therapeutics of neurodegenerative diseases.

Age has a powerful effect on enhanced susceptibility to neurodegenerative diseases, including susceptibility to stroke and cognitive impairment (CI) even in optimally healthy individuals. We critically evaluated the notion that oxidative stress increases in aging brain. Rigorous studies show logarithmic age-dependent increases in oxidized proteins and oxidized DNA lesions. Decreased activity of antioxidant protective enzymes does not account for the observed increases. The reactivity of the lipid oxidation product 4-hydroxy-2-nonenal (HNE) with key mitochondria enzymes may be important in the age-dependent loss in energy generation and enhanced susceptibility of neurons to apoptosis. Age-dependent enhanced neuroinflammatory processes may play an important role in toxin generation that causes death or dysfunction of neurons in neurodegenerative diseases. Non-steroidal anti-inflammatory drugs (NSAIDs) show significant promise. Vitamin E supplementation did not show major beneficial effect on cognitive functions. Major clinical trials for Alzheimers disease (AD) involving cycloxygenase-II (COX II) inhibitors and amyloid-beta vaccination have been discontinued. Novel therapeutics based on blocking neuron damaging neuroinflammatory processes show great promise for abating dementia progression although they have yet to make it to clinical practice.

Hydroxyl Free Radical Adduct of Deoxyguanosine: Sensitive Detection and Mechanisms of Formation

DNA or 2-deoxyguanosine reacts with hydroxyl free radical to form 8-hydroxy-deoxyguanosine (8-OH-dG). We found that 8-OH-dG can be effectively separated from deoxyguanosine by high pressure liquid chromatography and very sensitively detected using electrochemical detection. The sensitivity of electrochemical detection is about one-thousand fold enhanced over optical detection. Utilizing deoxyguanosine in bicarbonate buffer it was found that ferrous ion, but not ferric ion, was effective in forming 8-OH-dG. The hydroxyl free radical scavenging agents, thiourea and ethanol, were very effective in quenching Fe(11) mediated 8-OH-dG formation, but superoxide dismutase had very little effect.

Oxygen free radical involvement in ischemia and reperfusion injury to brain.

Evidence is presented which implicates increased oxygen free radicals during ischemia reperfusion of gerbil brain. Salicylate, which reacts with hydroxyl free radicals to yield dihydroxybenzoic acid (DHBA), was used as an in vivo trap. Brain ischemia for at least 5 min followed by reperfusion yielded significantly increased brain DHBA. Without reperfusion or with only 2 min of ischemia and then reperfusion, the production of DHBA was not increased. Increased levels of DHBA in brain correlated with ischemia reperfusion-mediated behavioral modification of gerbils, but salicylate administration did not protect against the behavior changes.

Sensitive assay of hydroxyl free radical formation utilizing high pressure liquid chromatography with electrochemical detection of phenol and salicylate hydroxylation products

Evidence is presented for a sensitive method useful for the detection of hydroxyl free radical generation in various systems. The methodology employs high pressure liquid chromatography with electrochemical detection (LCED) for the quantification and identification of the hydroxylation products from the reaction of OH with both phenol and salicylate. A detection limit of less than 1 pmol for the hydroxylation products has been achieved with electrochemical detector responses linear over at least three orders of magnitude. Detection and quantitation of the hydroxylation products obtained and formed during OH generation from biologically meaningful systems have been demonstrated. The three systems utilized were ADP/FE(II)/H2O/, hypoxanthine/xanthine oxidase plus chelated iron, and UV photolysis of H2O2.

p38 kinase is activated in the Alzheimer's disease brain.

Abstract: The p38 mitogen‐activated protein kinase is a stress‐activated enzyme responsible for transducing inflammatory signals and initiating apoptosis. In the Alzheimers disease (AD) brain, increased levels of phosphorylated (active) p38 were detected relative to age‐matched normal brain. Intense phospho‐p38 immunoreactivity was associated with neuritic plaques, neuropil threads, and neurofibrillary tangle‐bearing neurons. The antibody against phosphorylated p38 recognized many of the same structures as an antibody against aberrantly phosphorylated, paired helical filament (PHF) tau, although PHF‐positive tau did not cross‐react with the phospho‐p38 antibody. These findings suggest a neuroinflammatory mechanism in the AD brain, in which aberrant protein phosphorylation affects signal transduction elements, including the p38 kinase cascade, as well as cytoskeletal components.

Regional lipid peroxidation in rat brain in vitro: possible role of endogenous iron.

Lipoperoxidative capacity of various brain areas of aging rats was examined in vitro using the thiobarbituric acid test. Significant regional differences in the generation of lipid peroxides were found in freshly prepared homogenates from different areas of brain incubated under air. Incubation under oxygen resulted in marked stimulation of lipid peroxidation, with highest increases in hypothalamus (144%). Addition of exogenous Fe2+ and ascorbic acid resulted in stimulation of lipid peroxidation ranging from 10-fold in cortex to 20-fold in hypothalamus homogenates during incubation in air. A linear relationship was found between endogenous iron content in brain regions and their ability to produce lipid peroxides in vitro under oxygen for all areas except striatum. Several iron chelating agents effectively inhibited lipid peroxidation under hyperbaric oxygen whereas oxygenfree radical scavengers, as well as catalase and superoxide dismutase were not effective. It is concluded that regional differences in lipoperoxidative capacity of brain areas in vitro are in part governed by local endogenous iron content and may indicate regional susceptibility to oxidative damage.

Temporal patterns of cytokine and apoptosis-related gene expression in spinal cords of the G93A-SOD1 mouse model of amyotrophic lateral sclerosis.

Familial amyotrophic lateral sclerosis (FALS) is often caused by gain-of-function mutations in Cu,Zn-superoxide dismutase (SOD1). Multiprobe ribonuclease protection assays (RPAs) were used to investigate expression of 36 different cytokines and apoptosis-related genes in spinal cords of mice that ubiquitously express human SOD1 bearing a glycine (r) alanine substitution at residue 93 (G93A-SOD1). Mice were studied at late presymptomatic stage (80 days), and at 120 days when the animals experience severe hindlimb paralysis and accumulation of oxidatively modified proteins. Spinal cord tissue from G93A-SOD1 mice expressed a selective subset of macrophage-typical cytokines (monokines) including interleukin (IL)1alpha, IL1beta and IL1RA at 80 days increasing by 120 days. Contrastingly, T-cell derived cytokines (lymphokines) including IL2, IL3 and IL4 were detected at low levels in non-transgenic mice but these were not elevated in G93A-SOD1 mice even at 120 days. Apoptosis-related genes were generally unaffected at 80 days but multiple caspases and death receptor components were up-regulated at 120 days; the only exceptions being FADD and the tumor necrosis factor (TNF)alpha receptor p55 which was up-regulated at 80 days and increased further at 120 days. These data indicate that in the G93A-SOD1 mouse: (i) cytokine expression changes precede bulk protein oxidation and apoptosis gene expression; (ii) lymphocyte contributions to cytokine expression in FALS are likely minor; and (iii) TNFalpha and its receptors may link inflammation to apoptosis in ALS.

Message and protein-level elevation of tumor necrosis factor α (TNFα) and TNFα-modulating cytokines in spinal cords of the G93A-SOD1 mouse model for amyotrophic lateral sclerosis

Recent data indicate that certain pro-inflammatory cytokines are transcriptionally upregulated in the spinal cords of G93A-SOD1 mice, a model of amyotrophic lateral sclerosis (ALS). We previously showed that the receptor for tumor necrosis factor alpha (TNF-R1) was notably elevated at late presymptomatic as well as symptomatic phases of disease (J. Neurochem. 82 (2002) 365). We now extend these findings by showing that message for TNFalpha, as well as mRNA for interferon gamma (IFNgamma) and transforming growth factor beta1/2 (TGFbeta1, TGFbeta2), is simultaneously increased. Furthermore, TNFalpha protein is significantly increased in G93A-SOD1 mouse spinal cords, as are protein levels for interleukin-6 (IL6), IFNgamma, and the chemokines RANTES (CCL5) and KC. The interaction of TNFalpha, IL6, and IFNgamma proteins was modeled in vitro using Walker EOC-20 murine microglia with nitrite (NO(2)(-)) efflux as a quantitative index of cell response. TNFalpha alone caused robust NO(2)(-) flux, while IL6 had a lesser effect and neither IFNgamma nor IL1beta was active when applied singly. The TNFalpha stimulus was potently magnified in the presence of IL6 or IFNgamma. When applied in combination at very low concentrations, IFNgamma co-synergized with IL6 to produce a multiplicative increase in NO(2)(-) after stimulation with TNFalpha. Taken together, these data suggest that modest increases in multiple synergistic cytokines could produce a disproportionately severe activation of microglia within the degenerating spinal cord. Our data support a model wherein TNFalpha acts as a principal driver for neuroinflammation, while several co-stimulating cytokines and chemokines act to potentiate the TNFalpha effects.