Reactive Oxygen Species and Protein Oxidation in Aging: A Look Back, A Look Ahead

Archives of Biochemistry and BiophysicsVol. 397, No. 2, January 15, pp. 377–383, 2002doi:10.1006/abbi.2001.2630, available online at http://www.idealibrary.com on

MINIREVIEW

Reactive Oxygen Species and Protein Oxidation in Aging:A Look Back, A Look Ahead

Kenneth Hensley1 and Robert A. FloydFree Radical Biology and Aging Research Program, Oklahoma Medical Research Foundation,Oklahoma City, Oklahoma 73104

Received September 5, 2001, and in revised form October 8, 2001; published online December 13, 2001

The existence of free radicals, as chemical entities,was inferred 100 years ago but not universally ac-cepted for some 30–40 years. The existence and impor-tance of free radicals in biological systems was notrecognized until the mid 1950s, by a small number ofvisionary scientists who can be credited with found-ing the field of reactive oxygen biochemistry. For mostof the remaining 20th century, reactive oxygen species(ROS) were considered a type of biochemical “rustingagent” that caused stochastic tissue damage and dis-ease. As we enter the 21st century, reactive oxygenbiochemistry is maturing as a discipline and establish-ing its importance among the biomedical sciences. It isnow recognized that virtually every disease state in-volves some degree of oxidative stress. Moreover, weare now beginning to recognize that ROS are pro-duced in a well-regulated manner to help maintainhomeostasis on the cellular level in normal, healthytissue. This review summarizes the history of reactiveoxygen biochemistry, outlining major paradigm shiftsthat the field has undergone and continues to experi-ence. The contributions of Earl Stadtman to the recenthistory of the field (1980-present) are especially high-lighted. The role of ROS in signal transduction is pre-sented in some detail as central to the latest paradigmshift. Emerging technologies, particularly proteomictechnologies, are discussed that will facilitate furtherevolution in the field of reactive oxygen biochemistry.© 2001 Elsevier Science

Key Words: reactive oxygen; free radical history; sig-nal transduction; protein oxidation.

1 To whom correspondence and reprint requests should be ad-
dressed. Fax: (405) 271-1795. E-mail: [email protected].
0003-9861/01 $35.00© 2001 Elsevier ScienceAll rights reserved.

FENTON’S REACTION AND THE ORIGIN OF FREERADICAL RESEARCH

Gaseous oxygen (O2) has been recognized as both ananimating and a toxic principle since Priestley,Scheele, and Lavosier first isolated the gas and notedits curious behavior. Periodically during the 19th andearly 20th Century, some notice was taken of the toxiceffects produced by pure oxygen, particularly seizuresand lung injury (1). The first discovery of solution-phase ROS chemistry is universally accredited toHenry John Horstman Fenton (1854–1929). Koppenol(2) relates the serendipitous discovery of this chemistryby Fenton when he was a young student at Cambridge.Apparently, a fellow student was mixing chemicals atrandom and found that hydrogen peroxide combinedwith tartaric acid and ferrous (but not ferric) salts toyield a beautiful violet solution. In 1876, Fenton re-counted this experiment in a letter to Chemical News,as a possible diagnostic test for tartrate (3). The iron/peroxide chemistry fascinated Fenton for decades,leading him to identify tartrate hydroxylation productsand glycolic aldehyde in 1893 (4), which is generallytaken as the date of his seminal discovery. Fenton’srecognition that ROS could introduce reactive diols andcarbonyl groups into biomolecules would later proveimmensely significant to protein chemists. Fenton la-bored the rest of his life to discover the mechanism ofhis reaction, without success. It should be rememberedthat Thompson had not proven the existence of elec-trons until 1897, and Bohr’s atomic model was onlyproposed in 1911. The first tentative report of a freeradical per se is attributed to Moses Gomberg (1900)who synthesized triphenylmethyl (5). Many chemistsprior to 1930 doubted the veracity of organic free rad-
icals. Sadly, though Fenton’s work would massively
377

378 HENSLEY AND FLOYD

influence free radical biochemistry, he died two yearsbefore Haber and Willstatter first suggested the exis-tence of hydroxyl radical (•OH) (6). In 1934, Haber andWeiss reported the metal-catalyzed peroxide decompo-sition reaction, essentially as it is now understood (7).Haber and Weiss did not mention Fenton, who waslikely unknown to the German chemists. Koppenolsuggests (2) that hydroxyl radicals were not associatedwith Fenton’s organic transformations until 1946–1950, shortly before the birth of reactive oxygen bio-chemistry.

THE BIRTH OF REACTIVE OXYGEN BIOCHEMISTRY

It is often difficult to trace the origins of ideas thatevolve at the boundaries of disparate scientific disci-plines. Such is the case for the discovery of free radicalsin biological systems. By the middle 1950s, free radi-cals were accepted chemical entities outside of biology.To this point, most scientists who studied free radicalswere interested in novel organic syntheses. Global war-time necessities stimulated the organic chemical dis-coveries; for instance, free radical chain reactions wereused to synthesize polyethylene and were investigatedas routes to synthetic rubber. Antioxidants, accord-ingly, were understood as chain-breaking agents andinhibitors of free radical syntheses. In 1955, Wallingpublished a classic text that summarized contempo-rary knowledge of solution-phase free radical syntheticchemistry (8). If free radicals were studied in a biolog-ical context, this work was mostly restricted to under-standing the autoxidation of fats in foods and methodsto prevent spoilage (another significant necessity to thewar effort). Vitamin E (a-tocopherol) had been discov-ered in 1922 as a substance whose absence causedmammalian infertility (9) and subsequently was recog-nized to act as an inhibitor of fatty acid oxidation instored foodstuffs (10).

World War II events led directly to the birth of freeradical biochemistry. Radiation poisoning and radia-tion-induced mutations had been painfully demon-strated to a horrified world, and much of the post WWIIscientific effort was aimed at understanding radiationsickness. In a seminal paper published in 1954, Re-becca Gershman and Dan Gilbert first speculated thatthe lethal effects of ionizing radiation might be as-cribed to oxygen free radicals (11). A bold work ofintellectual synthesis and logical extrapolation, thismanuscript probably represents the birth of “reactiveoxygen biochemistry” as a discreet discipline and needsto be considered in light of its contemporary scientificcontext. At this time, X-irradiation was known to gen-erate radicals in solution but oxyradicals were other-wise unknown to biochemistry. Using the newly in-vented technique of electron paramagnetic resonance
(EPR), Barry Commoner had just shown that germi-
nating barley contained free radicals; this was the firstproof that free radicals exist in a biological organism(12). Commoner noted these radicals were either pro-teinaceous or associated with biopolymers such as mel-anin (12). As early as 1951, contemporaries such asLeonor Michaelis had recognized the existence of pro-tein-bound flavin semiquinone radicals by means ofoptical spectroscopy (13), but did not seriously considerthe existence and significance of diffusible oxyradicalsas has sometimes been claimed. It is historically sig-nificant that these early experimentalists insistedupon the protein-bound nature of biological free radi-cals; almost thirty more years would pass before solu-tion-phase oxyradicals would be widely accepted ascredible biochemical entities. Among other luminariesof the day, Britton Chance was perhaps the only majorinvestigator to empirically consider diffusible ROSwhen he studied the ability of catalase to reduce H2O2

(14). As early as 1956, Chance notes incidentally thatH2O2 is formed by respiring mitochondria (15).

Considering this early literature and noticing simi-larities in the type of tissue damage caused by bothhyperoxia and irradiation, Gershman and Gilbert triedtwo experiments. In the first they attempted to protectmice against hyperoxia by administering compounds,such as the antioxidant propyl gallate, that wereknown radioprotectants. They found that the sameagents were protective in both paradigms. In the sec-ond experiment, Gershman and Gilbert tested the ef-fect of sublethal radiation exposure on subsequent sen-sitivity to hyperoxia. This study found that the lethaleffects of the two challenges were synergistic. At theconclusion of their paper, Gershman and Gilbert write“. . . it would appear that irradiation and oxygen poi-soning produce some of their lethal effects through atleast one common mechanism, possibly that of the for-mation of oxidizing free radicals” (11). This statementwould not be widely accepted as an important biomed-ical insight for two decades, largely owing to the nearimpossible task of proving the existence of oxyradicalswithin biological fluids. Indeed, biological free radicalsremained the pervue of physicists and physical chem-ists.

Following close behind the Gershman manuscriptcame Denham Harman’s classical hypothesis paper,which argued the “Free Radical Theory of Aging” (16).Harman argued that the tempo of aging could be re-lated to steady-state levels of radicals produced, prob-ably, by metals or by respiratory complexes (16). BothHarman and Gilbert suggested that cellular thiolsmight be particular targets of radiation/radicals andthat supplementation of these pools could prove bene-ficial. This was well before biochemists had an under-standing of thiol-dependent redox regulation or therole of thiols in cellular signal transduction (discussed
below). Not surprisingly, both the Harman paper and

379ROS AND PROTEIN OXIDATION IN AGING

the Gershman/Gilbert paper (and many other studiesat this time) were written under the aegis of the AtomicEnergy Commission (11, 16).

The pioneering work of Gilbert, Chance, Commoner,Harman and their colleagues was performed at a timewhen biologists had only a vague notion of the struc-tural basis for biochemistry. To put this period in per-spective, Watson and Crick had published the DNAstructure in 1953. The genetic code was not completelybroken for another 15 years. John Kendrew deter-mined the first X-ray crystal structure of a protein in1957. Singer and Nicolson’s fluid mosaic model for cellmembranes was not published until 1972. Thus, thevisionary thinking of these “founders” of reactive oxy-gen biochemistry should be properly appreciated.Nonetheless, the early work inadvertently establisheda dogma that would persist to this day and which,perhaps, has held the field back in some respects. Thiswas the idea that oxidative stress is a purely stochasticevent, causing accumulated (but random) damage toproteins, lipids, and nucleic acids. Thus the cell wasconsidered a homogeneous environment within whichoccurred solution-phase chemistry, and oxidativestress was thought of as a type of gross cellular rusting.

THE VALIDATION OF REACTIVE OXYGENBIOCHEMISTRY: 1950–1980

Decades of research were devoted to determiningwhether free radicals are important biochemical enti-ties and what at relationship, if any, random oxidativedamage might have to disease. Lipid peroxidation wasgenerally studied more than protein or DNA oxidation,perhaps owing to the historical precedence of studieson lipid autoxidation in foodstuffs, and perhaps be-cause to the relative ease of measuring lipid oxidation(albeit by relatively nonspecific techniques such as theoptical assessment of conjugated dienes or by thiobar-bituric acid reactivity). It was realized that lipid oxi-dation occurs in many disease states and that thecytochrome P-450 system could convert halogenatedhydrocarbons into carbon-centered radicals capable ofinitiating lipid peroxidation chains. EPR methods wereused more and more to study these processes, and weremuch advanced by Janzen’s introduction of the freeradical spin trapping technique in 1968, which allowedthe detection of many unstable radicals at room tem-perature (17). While DNA was known to be sensitive tomutanizing effects of ionizing radiation, the role ofROS remained highly controversial. Facile techniquesfor measuring oxidized DNA in biological tissue werenot introduced until 1986, when electrochemical tech-niques were invented for the determination of 8-oxo-deoxyguanosine and hydroxylated salicylate (the latterbeing used as an exogenous trap for oxyradicals) (18).
Owing largely to the technical difficulty of measuring
radicals and oxidative damage, it was very difficult toprove that ROS could contribute meaningfully to dis-ease. Prior to 1970 most mainstream biochemistsdoubted that free radicals could contribute causally to“natural” disease processes. The main reason for thisskepticism was widespread doubt that oxyradicalscould exist as solution-phase entities, except perhapsunder conditions of extremely intense ionizing radia-tion.

This reticence was somewhat overcome in 1969 whenIrwin Fridovich and Joe McCord discovered Cu, Zn-superoxide dismutase (SOD1), an enzyme whose solefunction apparently was to detoxify the superoxide rad-ical anion (19–22). The ubiquitous presence of SOD1was taken as “proof” that diffusible biological oxyradi-cals must exist commonly in the solution phase, andnot merely as rare protein-bound oxidant species. Thisparadigm shift cannot be overemphasized. In a reviewon the SOD1 discovery, Bannister emphasizes that“. . . The discovery of superoxide dismutase had themakings of a Kuhnian scientific revolution” (22). Asecond, overlooked consequence of the McCord/Fridov-ich discovery was a refocusing of the field on proteinsas generators, consumers, and targets of oxyradicals.This refocus attracted the attention not only of physi-cal chemists but also that of mainstream protein bio-chemists, who began to study ROS biochemistry inmicrostructural detail to a degree heretofore unprece-dented.

PROTEIN OXIDATION AS PART OF THE HUMANAGING PROCESS: AN HOMAGE TO EARL STADTMAN

By 1980, oxyradicals were accepted biological enti-ties though their significance to aging and disease re-mained generally unappreciated. Tools such as EPRspectrometry existed to seriously study free radicalbiochemistry, though the field would long continue tobe plagued by shortcomings in bioanalytical methodol-ogies. Radical biochemistry was largely restricted tothe study of fulminating lipid peroxidation in livermicrosomes or in liver undergoing xenobiotic stress.This focus was due largely to (1) historical precedentand relative accessibility of tools such as spin trappingor optical methods for measuring these processes; and(2) an implicit dogma that oxyradical stress was sto-chastic and universally detrimental. There was ade-quate reason to extend these intellectual boundaries.Protein had long been known to experience nonenzy-matic glycation, concomitant with formation of protein-bound diols similar to the entities formed in Fenton’sreactions (discussed above). Pathologists had long rec-ognized that polymeric pigments (lipofuscin and thelike) collect in an age-dependent manner, particularlyin brain; and Commoner had shown in 1954 that such
structures form stable free radicals. Between 1970–


1985, David Gershon had been carefully documentingthe age-related accumulation of inactive enzymes (par-ticularly aldolase) that were very possibly damaged byoxidative events though this was never specifically ad-dressed (22–24). However, prior to 1980 no conceptualframework existed to integrate protein oxidation intometabolic regulation or to the aging process.

Around 1980, Earl Stadtman and his students beganto seriously investigate the nature and consequences ofprotein oxidation in vitro and in vivo. Stadtman hadextensively investigated glutamine synthetase (GS), akey metabolic enzyme that converts glutamine into anitrogen source for pathways leading to biosynthesis ofamino acids, nucleic acids and complex polysaccha-rides. GS is a complex metalloenzyme with multiplelevels of allosteric regulation. In 1980–81 Earl Stadt-man, Cynthia Oliver, And Rod Levine published a se-ries of reports detailing the coupling between metal-catalyzed oxidation of GS to proteolytic degradation ofthe damaged enzyme (25, 26). This oxidation was verysite-specific, affecting one histidine and one arginineper GS subunit near the divalent metal binding sites(27, 28). This was starkly contrasted to the case ofionizing radiation, which tended to damage proteins ina more stochastic manner (29, 30). Only recently it hasbeen recognized, also by the Stadtman group, thatmetal-catalyzed oxidation is very product specific: themajority of these oxidized amino acid side chains canbe specifically attributed to glutamic or aminoadipicsemialdehydes (31). Building on this early work, Stadt-man’s group found that many (but by no means all) keymetabolic enzymes are readily inactivated by what hetermed “mixed-function” oxidation (MFO) systems in-volving O2, Fe or Cu, and an auxiliary reducing agent(32, 33). These findings lead to the “biomarker” conceptwherein loss of GS activity, along with other oxidation-sensitive activities, would often be taken as strongevidence that a tissue had experienced oxidative stress(34).

Stadtman’s group developed or popularized tech-niques for measuring protein carbonyl groups as indi-ces of oxidative damage, and applied these techniquesto the study of protein oxidation in aging tissue. Inlandmark papers published in 1987 and 1991, Stadt-man published the first detailed determinations of pro-tein oxidation in models of human aging. The firststudy (35) investigated erythrocytes of different ages,isolated from the same donor, and found that erythro-cyte proteins from older cells were more amenable toreductive alkylation using [H3]NaBH4 as a probe.Moreover, protein carbonyl levels in fibroblasts werefound to increase in a nonlinear fashion with age, usingdinitrophenylhydrazine as a probe for carbonyl targetsof reductive alkylation (35). Most intriguingly, proteincarbonyls were especially elevated in cultures from
patients with the rapid-aging syndrome progeria or
Werner’s syndrome (a defect in copper homeostasis)(35). Protein oxidation and enzyme biomarker activitywere highly correlated. In the 1991 manuscript, Stadt-man and colleagues show that protein oxidation in-creases exponentially in aging human brain tissue andis quantitatively elevated in affected regions of Alzhei-mer’s diseased brain tissue than in age-matched, nor-mal tissue (36). As much as 30% of the protein in agedcells was subsequently estimated to be carbonylated,verses perhaps 10% in youthful tissue (37). The re-search into Alzheimer’s disease has been replicatedand extended to include other oxidative post-transla-tional protein modifications, including oxidative ty-rosine crosslinks and nitrated tyrosine residues (38).The quantitative analyses performed by Stadtman andhis colleagues in the 1980s provided the strongest ev-idence, to that time, in support of the Gilbert andHarman hypotheses regarding oxidative stress as asignificant variable in human aging.

Earl Stadtman’s work has influenced the develop-ment of free radical biochemistry in a subtle but pro-found way. By seeking to understand protein oxidationon a very precise, structural and quantitative level,Stadtman’s work has revised the way we think aboutoxidative damage. In many cases we now seek to iden-tify the exact number, location and type of oxidativelymodified protein side chains within a particular bi-omolecule; we also wish to define tissue location, cellu-lar and even subcellular localization of the oxidativealteration. We have come to understand that age-re-lated oxidative stress is not so stochastic a process ashad once been thought. On a different level, Stadt-man’s application of simple but elegant bioanalyticaltechniques to human tissue has made reactive oxygenbiochemistry seem less academic and prompted a veryserious consideration of oxidative stress as a patholog-ical factor.

REACTIVE OXYGEN BIOCHEMISTRY AT THE STARTOF THE 21st CENTURY

Free radical research remains a rapidly growing andevolving discipline. A casual review of the PubMeddatabase (http://www.ncbi.nlm.nih.gov) of biomedicalpublications shows the fulminating condition of thefield. The number of publications indexed under thesearch term “free radical” has been doubling approxi-mately every 7 years since 1970 (data not illustrated).By comparison, the doubling rate for articles search-able under “hypertension” has been approximately 15years. While this is admittedly a crude indicator, it isclear that reactive oxygen biochemistry is no longer aminor subdiscipline. It is also clear that the reactiveoxygen researcher will become increasingly challengedto stay abreast of the literature and to identify major
trends in the field. In keeping with the historical pat-


terns of reactive oxygen research, we should anticipateparadigm shifts and guard against becoming two com-fortable with established dogma. One area in whichthis caveat may become germane is the topic of reactiveoxygen species (ROS) as signal transduction modula-tors, discussed in the final section of this review.

REACTIVE OXYGEN SPECIES AS NECESSARY AGENTSIN THE MAINTENANCE OF BIOLOGICALHOMEOSTASIS: THE LATEST PARADIGM SHIFT

ROS traditionally have been considered as malignagents: unavoidable side effects of life in an aerobicenvironment. Even when oxidative damage was recog-nized as rather nonstochastic, the ROS were consid-ered almost exclusively from the toxicological perspec-tive. This paradigm was abruptly challenged with thediscovery of nitric oxide (•NO) in 1987, and its identi-fication as the long-sought “endothelium derived relax-ing factor” (39). Within a short space of time it wasrecognized that a specific oxyradical was obligatory forproper vascular function. Not only could this radical besynthesized and degraded in a very controlled fashion,by specialized biochemical components; it was evenfound to bind hemoglobin as part of a previously un-known vasoregulatory mechanism (40). These findingswere tremendously surprising; many accomplished re-searchers were left wondering how a molecule as sim-ple, ubiquitous and necessary as NO could have re-mained undiscovered for so long.

Largely resulting from NO discovery, a new empha-sis has emerged on uncovering roles of biological ROSin cellular signal transduction. During the past decadeit has become clear that ROS activate multiple tran-scription factors including NFkB, which was one of thefirst ROS-sensitive transcription factors and so hasreceived perhaps a disproportionate amount of atten-tion (for reviews see 42, 43). NFkB and other ROS-sensitive transcription factors arbitrate inflammatoryreactions; growth and differentiation pathways; andapoptotic decisions. The pathological activation of suchsignaling systems is therefore a subject of great impor-tance. Early work showed that exogenous oxidants,especially H2O2, could stimulate nuclear localization ofNFkB when applied to cell culture (41). Some studieshave tracked the activation pathway past the level ofgene transcription. In our laboratory, for instance, wefind that reasonably low concentrations of exogenousH2O2 (100 mM exogenous bolus) can cause interleukintranscription in cultured primary astrocytes (44).Smaller doses of peroxide, perhaps as low as 10–20mM, can elicit measurable changes in phosphorylationof specific regulatory proteins including protein kinaseB (AKT) (44). Interleukin transcription appears to besecondary to the activation of the p38-mitogen acti-
vated protein kinase (p38-MAPK) cascade, and is re-
produced by inhibitors of mitochondrial respirationthat are known to stimulate intracellular peroxide re-lease (45, 46). Thus, both intracellular and extracellu-lar sources of ROS are capable of modulating geneexpression even at ROS tensions that are not likely tocause gross cellular disruption. In the case of oxida-tively challenged astrocyte culture, at least, H2O2 canbe said to act as a functional surrogate for authenticIL1b; that is to say, many of the cellular responses toperoxide and IL1b are indistinguishable (44).

A detailed review of signal transduction is outsidethe scope of the present discourse. It is however impor-tant to note several themes that are currently emerg-ing. It has become clear that most cell types are capa-ble of synthesizing superoxide and/or H2O2 “on de-mand” in response to specific stimuli. In 1995, TorenFinkel’s group first presented convincing evidence thatnonphagocytic cells can synthesize H2O2 in response toa receptor-ligand interaction (47). Specifically, theseresearchers showed that vascular smooth muscle cellstreated with platelet-derived growth factor (PDGF)synthesize H2O2 concomitant with increases in proteintyrosine phosphorylation (47). Both peroxide genera-tion and tyrosine phosphorylation were blocked bycatalase and the thiol antioxidant/reducing agent N-acetyl cysteine (47). Subsequently, these observationswere replicated in variety of cell types exposed to spe-cific growth factors or cytokines (46, 48). For example,in our laboratory we have shown that astrocytes ex-posed to IL1b release measurable quantities of perox-ide within several minutes of cytokine application andthis phenomenon perfectly correlates with measurablealteration in protein phosphorylation dynamics (45).The detection of ROS in living cells has been somethingof a technical coup, and is generally done using laserconfocal imaging or fluorescence-assisted cell sorters todetect peroxide-dependent oxidation of fluorogenicdyes. Thus, new technical advances have once moreopened new lines of inquiry into biological ROS.

Sources and targets of “second messenger” ROS arestill being identified. One major source of de novo per-oxide is the membrane-bound NADPH oxidase. Thisenzyme complex assembles rapidly at the plasmamembrane upon appropriate stimulation and probablyrepresents one major source of ROS used in the prop-agation of intracellular messages (reviewed in 49). TheNADPH oxidase is an obvious candidate for a peroxidegenerator because phagocytic cells have long beenknown to utilize such a system for host-defense pur-poses. We have argued (44, 46) that mitochondria mayalso act as signal transducing elements. Several stud-ies suggest that mitochondrial alterations occur duringthe immediate phase of ligand-mediated signal trans-duction. In L929 fibroblasts, for instance, mitochon-drial superoxide production reportedly increases
within minutes of TNFa application and transmem-


brane potential (DC) decreases within 2–4 h (51, 52).The possible role of mitochondria as signal transduc-tion modulators deserves closer attention in futurestudies.

Comparably more is known about the targets of sig-nal-transducing ROS. Our current understanding ofsignal transduction largely corroborates Dan Gilbert’sprescient comments about the importance of cellularthiols as oxidative response elements. As might beanticipated from organic chemical precepts, nucleo-philic cysteine side chains are very susceptible to reac-tion with both H2O2 and NO (reviewed in 44). Thesereactions are encouraged by structural elements thatshift the thiol equilibrium toward the thiolate anion,and increase the nucleophilicity of the sulfur. Thus,protein tyrosine phosphatases are particularly sensi-tive to inactivation in response to H2O2 (44, 52–54).This has been demonstrated in vitro (52) and in vivo(46, 53, 54). Moreover Sue Goo Rhee has shown re-cently that EGF, which is known to elevate intracellu-lar ROS (48), causes immediate loss of nucleophilicityin protein phosphatase 1B (PTP-1B) during stimula-tion of cultured A431 epidermoid cells (54). Proteinphosphatase inactivation is generally reversible in thepresence of glutathione and thioredoxin enzymes (52)and is very specific. When transition metals are se-questered, as in most healthy cells, hydroxyl radicalsare precluded and H2O2 is expected to react with veryfew substrates except activated thiols. The inactivationof protein phosphatases is tantamount to the activa-tion of the cognate kinases. In the presence of a givenlevel of receptor activation, an intracellular proteinkinase cascade will always operate at higher efficiencywhen the antagonistic phosphatase elements are deac-tivated.

Our emerging knowledge of ROS in signal transduc-tion represents the latest paradigm shift in reactiveoxygen biochemistry. We have come to realize thatROS may engender cellular stress but not necessarilythrough irreversible and stochastic processes. IndeedROS appear to have been co-opted by evolution to servenecessary and useful purposes in the maintenance ofcellular homeostasis and in the communication of thecell with its external milieu. A complete biochemistryexists to deal with ROS, as toxic elements and as abenign factor. In diseased states, ROS biochemistry islikely to be perturbed or ill-controlled but in a way thatis amenable to pharmacological manipulation bymeans other than ingesting massive quantities of freeradical-scavenging antioxidants. The onus now restson structural and molecular biologists to ascertain thebiophysical basis for particular ROS-sensitive regula-tory events; and upon basic science researchers to pin-point the elements most likely to be affected in humandisease states. This task requires the development and
application of new technologies, such as mass spectro-
metric and proteomic strategies, which allow the studyof reversible protein oxidation at a very fine structurallevel. Such instrumental techniques have recentlybeen applied to the identification of S-nitrosylated pro-teins in whole mammalian brain (55) and to the local-ization of nitrosothiol moieties within the primary se-quence of isolated hemoglobin (56) and monomeric G-proteins (57).

SUMMARY

From the historical perspective it is apparent thatreactive oxygen biochemistry has evolved slowly andthrough a series of paradigm shifts, each driven to alarge extent by the judicious application of new bioana-lytical techniques at the level of protein chemistry. Wecan undoubtedly expect and happily anticipate morefrequent “reconceptions” in the postgenomic era, whenmolecular biology and proteomic technologies mergeand become standard biochemistry tools. In the imme-diate future, the most gratifying lines of inquiry willlikely be in the area of cellular signal transduction. Themost significant findings, as ever, will demand thecreative and possibly even iconoclastic application ofrigorous analytical methods to robust model systems.

ACKNOWLEDGMENTS

Some of the data discussed in this review was collected under theaegis of the National Institutes of Health (NS35747, EY06595,EY12190, RR15564, AG05119 and AG18945); the ALS Association(ALSA); the American Heart Association (AHA 0051176Z); and theOklahoma Center for Neuroscience (OCNS).

REFERENCES

1. Binger, C. A. L., Faulkner, J. M., and Moore, R. L. (1927) J. Exp.Med. 45, 849–845.

2. Koppenol, W. H. (1993) Free Rad. Biol. Med. 15, 645–651.3. Fenton, H. J. H. (1876) Chem. News 33, 190.4. Fenton, H. J. H. (1893) J. Chem. Soc. Proc. 10, 157–158.5. Gomberg, M. (1900) J. Am. Chem. Soc. 22, 757–771.6. Haber, F., and Willstatter, R. (1931) Chem. Ber. 64, 2844–2856.7. Haber, F., and Weiss, J. (1934) Proc. R. Soc. A147, 332–351.8. Walling, C. (1955) Free Radicals in Solution, Wiley, New York.9. Evans, H. M., and Bishop, K. S. (1922) Science 56, 650–651.

10. Golumbic, C., and Mattill, H. S. (1941) J. Am. Chem. Soc. 63,1279.

11. Gershman, R., Gilbert, D. L., Nye, S. W., Dwyer, P., and Fenn,W. O. (1954) Science 119, 623–626.

12. Commoner, B., Townsend, J., and Pake, G. E. (1954) Nature 174,689–691.

13. Michealis, L. (1951) in The Enzymes (Sumner, J. B., and Myr-back, K., Eds.), Academic Press, New York.

14. Chance, B. (1951) in The Enzymes (Sumner, J. B., and Myrback,K., Eds.), Academic Press, New York.

15. Chance, B., and Williams, G. R. (1956) Advan. Enzymol. Relat.Subj. Biochem. 17, 65–134.

16. Harman, D. (1956) J. Gerontol. 11, 298–300.


17. Janzen, E. G., and Blackburn, B. J. (1968) J. Am. Chem. Soc. 90,5909.

18. Floyd, R. A., Watson, J. J., Wong, P. K., Altmiller, D. H., andRickard, R. C. (1986) Free Rad. Res. Commun. 1, 163–172.

19. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chem. 244,6049–6055.

20. McCord, J. M., and Fridovich, I. (1977) in Superoxide and Su-peroxide Dismases (Michelson, A. M., McCord, J. M., and Fri-dovich, I., Eds.) pp. 1–10, Academic Press, London.

21. Bannister, W. H., and Bannister, J. F. (1988) Free Rad. Biol.Med. 5, 371–376.

22. Gershon, H., and Gershon, D. (1973) Proc. Natl. Acad. Sci. USA70, 909–913.

23. Dovrat, A., and Gershon, D. (1983) Biochim. Biophys. Acta 757,164–167.

24. Reznick, A. Z., Rosenfelder, L., Shpund, S., and Gershon, D.(1985) Proc. Natl. Acad. Sci. USA 82, 6114–6118.

25. Oliver, C. N. (1981) The Regulation of Glutamine SynthetaseDegradation, Ph.D. thesis, The Johns Hopkins University.

26. Levine, R. L., Oliver, C. N., Fulks, R. M., and Stadtman, E. R.(1981) Proc. Natl. Acad. Sci. USA 78, 2120–2124.

27. Farber, J. M., and Levine, R. L. (1986) J. Biol. Chem. 261,4574–4578.

28. Colombo, G., and Villafranca, J. J. (1986) J. Biol. Chem. 261,10587–10591.

29. Davies, K. J. A., Delsignore, M. E., and Lin, S. W. (1987) J. Biol.Chem. 262, 9902–9907.

30. Stadtman, E. R., and Oliver, C. N. (1991) J. Biol. Chem. 266,2005–2008.

31. Requena, J. R., Chao, C. C., Levine, R. L., and Stadtman, E. R.(2001) Proc. Natl. Acad. Sci. USA 98, 69–74.

32. Stadtman, E. R., and Wittenberger, M. E. (1985) Arch. Biochem.Biophys. 239, 379–387.

33. Fucci, L., Oliver, C. N., Coon, M. J., and Stadtman, E. R. (1983)Proc. Natl. Acad. Sci. USA 80, 1521–1525.

34. Hensley, K., Hall, N., Subramaniam, R., Cole, P., Harris, M.,Aksenov, M., Aksenova, M., Gabbita, S. P., Wu, J. F., Carney,J. M., Lovell, M., Markesbery, W. R., and Butterfield, D. A.(1995) J. Neurochem. 65, 2146–2156.

35. Oliver, C. N., Ahn, B-W., Moerman, E. J., Goldstein, S., andStadtman, E. R. (1987) J. Biol. Chem. 262, 5488–5491.

36. Smith, C. D., Carney, P. E., Starke-Reed, C. N., Oliver, E. R.,
Stadtman, E. R., Floyd, R. A., and Markesbery, W. R. (1991)Proc. Natl. Acad. Sci. USA 88, 10540–10543.
37. Starke-Reed, P. E., and Oliver, C. N. (1989) Arch. Biochem.Biophys. 275, 559–567.

38. Hensley, K., Maidt, M. L., Yu, Z. Q., Sang, H., Markesbery,W. R., and Floyd, R. A. (1998) J. Neurosci. 18, 8126–8132.

39. Palmer, R. M., Ferrige, A. G., and Moncada, S. (1987) Nature327, 524–526.

40. Gow, A. J., and Stamler, J. S. (1998) Nature 391, 169–173.41. Schreck, R., Rieber, P., and Baeuerle, P. A. (1991) EMBO J. 10,

2247–2258.42. Schreck, R., Albermann, K., and Baeuerle, P. A. (1992) Free Rad.

Res. Commun. 17, 221–237.43. Sen, C. K., and Packer, L. (1996) FASEB J. 10, 709–720.44. Gabbita, S. P., Robinson, K. A., Stewart, C. A., Floyd, R. A., and

Hensley, K. (2000) Arch. Biochem. Biophys. 376, 1–13.45. Robinson, K., Stewart, C. A., Pye, Q. N., Nguyen, X., Kenney, L.,

Salsman, S., Floyd, R. A., and Hensley, K. (1999) J. Neurosci.Res. 55, 724–732.

46. Robinson, K. A., Stewart, C. A., Pye, Q. N., Floyd, R. A., andHensley, K. (1999) Arch. Biochem. Biophys. 365, 211–215.

47. Sundaresan, M., Yu, Z. X., Ferrans, V. J., Irani, K., and Finkel,T. (1995) Science 270, 296–299.

48. Bae, Y. S., Kang, S. W., Seo, M. S., Baines, I. C., Tekle, E., Chock,P. B., and Rhee, S. G. (1997) J. Biol. Chem. 272, 217–221.

49. Irani, K., and Goldschmidt-Clermont, P. J. (1998) Biochem.Pharmacol. 55, 1339–1346.

50. Hennet, T., Richter, C., and Peterhans, E. (1993) Biochem. J.289, 587–592.

51. Polla, B. S., Jacquier-Sarlin, M. R., Kantengwa, S., Meriethoz,E., Hennet, T., Russo-Marie, F., and Cossirizza, A. (1996) FreeRad. Res. 25, 125–131.

52. Denu, J. M., and Tanner, K. G. (1998) Biochemistry 37, 5633–5642.

53. Barrett, W. C., DeGnore, J. P., Konig, S., Fales, H. M., Keng,Y. F., Zhang, Z. Y., Yim, M. B., and Chock, P. B. (1999) Biochem-istry 38, 6699–6705.

54. Lee, S. R., Kwon, K. S., Kim, S. R., and Rhee, S. G. (1998) J. Biol.Chem. 273, 15366–15372.

55. Jaffrey, S. R., Erdjument-Bromage, H., Ferris, C. D., Tempst, P.,and Snyder, S. H. (2001) Nat. Cell Biol. 3, 193–197.

56. Ferranti, P., Malorni, A., Mamone, G., Sannolo, N., and Marino,G. (1997) FEBS Lett. 400, 19–24.

57. Mirza, U. A., Chait, B. T., and Lander, H. M. (1995) J. Biol.Chem. 270, 17185–17188.

Documents

Reactive Oxygen Species and Protein Oxidation in Aging: A Look Back, A Look Ahead