You are on page 1of 35

PHYSIOLOGICAL REVIEWS

Vol. 78, No. 2, April 1998


Printed in U.S.A.

The Free Radical Theory of Aging Matures


KENNETH B. BECKMAN AND BRUCE N. AMES

Department of Molecular and Cell Biology, University of California, Berkeley, California

I. Introduction 548
II. An Overview of the Free Radical Theory of Aging 549
A. Origins of the free radical theory 549
B. Sources of oxidants 549
C. Targets of oxidants 552
D. Antioxidant defenses 553
E. Repair of oxidative damage 554

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


F. Synthesis: interaction of oxidant generation, oxidative damage, and repair 554
III. Refinements and Corollaries of the Free Radical Theory 554
A. Oxidants and evolutionary theories of aging 554
B. Oxidants and the somatic mutation theory of aging 555
C. Oxidants and mitochondrial theories of aging 556
IV. Oxidative Phenomenology: Age-Associated Trends 556
A. Accumulation of oxidative end products 556
B. Steady-state levels of oxidative modification 557
C. Oxidative depletion of biochemical pools 559
D. Age-associated trends in antioxidant defenses and repair 559
E. Age-associated trends in oxidant generation 560
V. Interspecies Comparisons 561
A. Oxidative damage and maximum life span potential 561
B. Antioxidant defenses and maximum life span potential 561
C. Generation of reactive oxygen species and maximum life span potential 562
VI. Dietary Restriction 562
VII. Manipulation of Rate of Living 563
VIII. Manipulation of Oxygen Tension 563
IX. Supplementation With Dietary Antioxidants 563
X. Administration of Pharmacological Antioxidants 564
XI. In Vitro Senescence and Oxidants 564
XII. Classical and Population Genetics 565
A. Caenorhabditis elegans genetics 565
B. Drosophila genetics 566
C. Rodent genetics 567
XIII. Molecular Genetics 567
A. Oxidants and nuclear somatic mutations 567
B. Oxidants and mitochondrial somatic mutations 567
XIV. Transgenic Organisms 568
A. Transgenic Drosophila 568
B. Transgenic mice 568
XV. Sporadic Degenerative Diseases 569
XVI. Inherited Degenerative Diseases 569
XVII. Epidemiology of Oxidants and Antioxidants 569
XVIII. Summary 570
A. Are oxidants responsible for aging? 570
B. Nitric oxide 571
C. Oxidants and apoptosis 571
D. Regulatory roles for oxidants and antioxidants: signal transduction 571

Beckman, Kenneth B., and Bruce N. Ames. The Free Radical Theory of Aging Matures. Physiol. Rev. 78: 547
581, 1998.The free radical theory of aging, conceived in 1956, has turned 40 and is rapidly attracting the interest
of the mainstream of biological research. From its origins in radiation biology, through a decade or so of dormancy
and two decades of steady phenomenological research, it has attracted an increasing number of scientists from an
expanding circle of fields. During the past decade, several lines of evidence have convinced a number of scientists

0031-9333/98 $15.00 Copyright q 1998 the American Physiological Society 547

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


548 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

that oxidants play an important role in aging. (For the sake of simplicity, we use the term oxidant to refer to all
reactive oxygen species, including O0
2 j, H2O2 , and jOH, even though the former often acts as a reductant and
produces oxidants indirectly.) The pace and scope of research in the last few years have been particularly impressive
and diverse. The only disadvantage of the current intellectual ferment is the difficulty in digesting the literature.
Therefore, we have systematically reviewed the status of the free radical theory, by categorizing the literature in
terms of the various types of experiments that have been performed. These include phenomenological measurements
of age-associated oxidative stress, interspecies comparisons, dietary restriction, the manipulation of metabolic
activity and oxygen tension, treatment with dietary and pharmacological antioxidants, in vitro senescence, classical
and population genetics, molecular genetics, transgenic organisms, the study of human diseases of aging, epidemio-
logical studies, and the ongoing elucidation of the role of active oxygen in biology.

I. INTRODUCTION that fruit flies and nematodes are amenable to the study
of aging. Also, medical researchers investigating human
The study of aging, by nature multidisciplinary, has diseases of aging, such as Alzheimers disease (AD) and
been characterized by a dizzying variety of theories, a inherited progerias, have overcome long-standing road-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


huge phenomenological literature, and the absence of blocks.
firmly established primary causes. The diverse life histo- It has been gratifying, therefore, that many of the
ries of animal species, which manifest aging in very differ- preliminary studies in what might be called molecular
ent ways, has been an obstacle to testing unified theories. gerontology lend credibility to the free radical theory.
For experimental gerontology to provide more than a cat- Results from disparate experimental systems have re-
alog of age-related changes, it has been necessary for cently shown that oxygen radicals play a role in degenera-
biologists to define the alterations that are common to tive senescence, and the pace of discoveries is quickening.
most old cells, tissues, and animals, while simultaneously The likely result of this collision of scientific approaches
respecting that there is not a single phenomenon of aging will be the unraveling of the physiological tangle of aging,
or a single cause. This has taken some time, and from the and it seems safe to say that one of the important knots
outside it may have appeared that the field has been mired will turn out to be oxidative stress.
in phenomenology. Perhaps for this reason, the study of However, there is a danger that in the excitement
aging was until recently avoided by molecular biologists, of theoretical confirmation, certain nuances are lost. For
who naturally favored clear-cut phenomena. As the molec- instance, the revelation that oxygen radicals may be in-
ular details of development, cancer, and immunology volved in neurodegeneration does not mean that oxidative
yielded to modern tools during the 1970s and 1980s, the stress determines life span. The free radical theory has
field of aging lagged, and mechanisms responsible for sought not only to explain the mechanisms of degenera-
aging failed to emerge. tive senescence, but it has also attempted to explain dif-
Throughout this time, though, there has never been ferences between species life spans in terms of oxidants.
a shortage of unified theories attempting to reduce aging So, although many recent studies indicate that oxygen
to something more tractable. In fact, gerontologists have radicals play some kind of role in aging, only a small
been prolific in this regard (83, 205). Whereas some re- number of these support the more ambitious version of
searchers have believed that a small number of random, the free radical theory. On the other hand, there is no
deleterious mechanisms could explain degenerative se- reason to cling to such a stringent version of the free
nescence, others have opted for theories of pro- radical theory, and it is becoming apparent that whether
grammed aging, in which senescence is the final destina- or not they determine life span, oxygen radicals are cer-
tion in a developmental pathway. In the course of these tainly important players in agings pathophysiology. In
debates, a number of scientists have rallied around a set other words, the scope of the free radical theory of aging
of ideas called the free radical theory of aging: loosely, the should include aging-associated oxidative stress in gen-
belief that damage by reactive oxygen species is critical in eral, rather than limiting itself to those oxidative events
determining life span. This theory inspired many experi- that may determine life span. In fact, many current articles
ments in which evidence of oxidative damage in aged indicate that such a blurring of distinctions has already
animals was sought. occurred and that as it is commonly used, free radical
In the last 10 years or so, the nature of aging research theory encompasses a broad set of ideas. Therefore, our
has changed dramatically; one might say that the field has first purpose is to delineate these different conceptions
entered early adulthood. The tools of molecular biology of the free radical theory, as a prerequisite to its critical
are now sophisticated and accessible enough that re- evaluation.
searchers within gerontology have adopted them. At the Because of the recent popularity of free radical re-
same time, molecular biologists situated on the edge of search, a large number of reviews have addressed various
aging research have made inroads and have discovered aspects of the interplay between oxidants and aging (6,

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 549

15, 18, 33, 51, 57, 58, 60, 65, 82, 85, 87, 103, 106, 123126, sized that traces of iron and other metals would catalyze
130, 161, 166, 196, 203, 227, 257, 273, 275, 288, 293, 297, oxidative reactions in vivo and that peroxidative chain
307, 312, 315, 335, 338, 348, 353, 357). Rather than merely reactions were possible, by analogy to the principles of
updating this literature, our aim is to provide a systematic in vitro polymer chemistry. All of these predictions have
categorization of the types of experiments that have been been confirmed during the past 40 years.
performed. The phenomenon and study of aging are in- The theory gained credibility with the identification in
credibly diverse, encompassing organisms from rotifers 1969 of the enzyme superoxide dismutase (SOD) (204),
to mammals and techniques from physiology to genetics. which provided the first compelling evidence of in vivo
Although it is precisely the broad sweep of evidence that generation of superoxide anion (O20j), and from the subse-
lends the free radical theory its appeal, the menagerie of quent elucidation of elaborate antioxidant defenses (351).
animals and techniques sometimes obscures the logic. By The use of SOD as a tool to locate subcellular sites of
breaking the literature down into smaller pieces (a prac- O20j generation led to a realization that buttressed the free
tice we find necessary ourselves), we hope to make it radical theory, namely, that mitochondria are a principal
easier for readers to judge the theory. Moreover, by impos- source of endogenous oxidants (37). Gerontologists had
ing a structure, we aim to highlight novel and definitive long observed that species with higher metabolic rates

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


approaches, because it is these that will replace the phe- have shorter maximum life span potential (MLSP); they
nomenology of past decades. age faster (268). In fact, it had been proposed at the turn
In this review, then, we briefly outline the evolution of the century that energy consumption per se was respon-
of the free radical theory and then delineate the different sible for senescence, a concept referred to as the rate of
areas of evidence. We focus on recent experiments and living hypothesis (246, 268, 295). The realization that en-
point to the areas that we feel are most likely to provide ergy consumption by mitochondria may result in O20j pro-
future insights. The way in which we have categorized duction linked the free radical theory and the rate of living
the literature is outlined in the table of contents (sects. theory irrevocably: a faster rate of respiration, associated
IV-XVII) and in Table 1. Although any such system is some- with a greater generation of oxygen radicals, hastens aging.
what arbitrary, we hope that ours will make it easier both By now, the two concepts have essentially merged.
to assimilate the existing literature and to envision future
experiments. In writing a review on as broad a topic as
the free radical theory, we have been forced to limit both B. Sources of Oxidants
the content and the number of references cited. Although
we have done our best to include recent work, omissions Ground-state diatomic oxygen (3Sg0O2 or more com-
were inevitable. We apologize to all authors whose work monly, O2), despite being a radical species and the most
we have not managed to include, and direct readers to important oxidant in aerobic organisms, is only sparingly
other recent reviews for material we have left out. reactive itself due to the fact that its two unpaired elec-
trons are located in different molecular orbitals and pos-
sess parallel spins. As a consequence, if O2 is simultane-
II. AN OVERVIEW OF THE FREE RADICAL
ously to accept two electrons, these must both possess
THEORY OF AGING
antiparallel spins relative to the unpaired electrons in O2 ,
a criterion which is not satisfied by a typical pair of elec-
A. Origins of the Free Radical Theory trons in atomic or molecular orbitals (which have oppo-
site spins according to the Pauli exclusion principle). As
In 1956, Denham Harman suggested that free radicals a result, O2 preferentially accepts electrons one at a time
produced during aerobic respiration cause cumulative ox- from other radicals (such as transition metals in certain
idative damage, resulting in aging and death. He noted valences). Thus, in vivo, typical two- or four-electron re-
parallels between the effects of aging and of ionizing radi- duction of O2 relies on coordinated, serial, enzyme-cata-
ation, including mutagenesis, cancer, and gross cellular lyzed one-electron reductions, and the enzymes that carry
damage (120, 128). At the time, it had recently been dis- these out typically possess active-site radical species such
covered that radiolysis of water generates hydroxyl radi- as iron. One- and two-electron reduction of O2 generates
cal (jOH) (319), and early experiments using paramag- O20j and hydrogen peroxide (H2O2), respectively, both of
netic resonance spectroscopy had identified the presence which are generated by numerous routes in vivo, as dis-
of jOH in living matter (45). Harman (120) therefore hy- cussed below. In the presence of free transition metals
pothesized that endogenous oxygen radical generation oc- (in particular iron and copper), O20j and H2O2 together
curs in vivo, as a by-product of enzymatic redox chemis- generate the extremely reactive hydroxyl radical (jOH).
try. He ventured that the enzymes involved would be those Ultimately, jOH is assumed to be the species responsible
involved in the direct utilization of molecular oxygen, for initiating the oxidative destruction of biomolecules.
particularly those containing iron. Finally, he hypothe- In addition to O20j, H2O2 , and jOH, two energetically ex-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


550 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

TABLE 1. The strengths and weaknesses of approaches to the testing of the free radical theory of aging
Experimental Approach
(Review Section) Strengths of the Approach Weaknesses of the Approach

Oxidative phenomenology Simplicity; large existing data set; elucidation of the Results are merely correlative (oxidative damage
(Sect. IV) basic biochemistry of oxidative stress; technical may be a consequence of aging); negative
foundation for other approaches; negative result results, considered uninteresting, may fail to
may be instructive appear in the literature
Interspecies comparisons Specific testable predictions; deviations from Logistical problems in animal handling;
(Sect. V) predictions may refine theory; use of different quantitative comparisons complicated by
species avoids conclusions that are species or qualitative interspecies differences in oxidative
strain specific; negative results may be instructive defenses and/or repair
Dietary restriction Very well established model of life span extension; Results are somewhat correlative (alterations in
(Sect. VI) straightforward methodology; probably relevant to oxidative stress may be consistent with but
human aging and cancer incidental to a more fundamental cellular
switch)
Manipulation of rate of A direct test of rate of living theory, with specific To date, limited to relatively simple organisms,
living testable predictions; straightforward methodology such as invertebrates

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


(Sect. VII)
Manipulation of oxygen A direct test of rate of living theory, with specific Results hard to interpret. Are positive results
concentration testable predictions; straightforward methodology (life span attenuation by hyperoxia, extension
(Sect. VIII) by hypoxia) relevant to normoxic aging? Not
applicable to most mammals
Supplementation with (Potentially) a test of free radical theory Results hard to interpret, due to unknown fates
dietary antioxidants of supplements, potential adverse effects,
(Sect. IX) complexity of overall antioxidant defenses
Administration of (Potentially) a test of free radical theory Results hard to interpret, due to unknown fates
pharmacological of supplements, potential adverse effects,
antioxidants complexity of overall antioxidant defenses
(Sect. X)
In vitro senescence A very well-established model system; relevance to Relevance to aging in vivo not yet clear
(Sect. XI) human cancer (and perhaps to degenerative
senescence in vivo)
Classical and population Awesome power of genetics; very well established Currently limited to invertebrate model
genetics methods linked to a vast body of species-specific organisms (yeast, C. elegans, Drosophila)
(Sect. XII) information; concurrent genome projects may
forge interspecies links between gerontogenes
Molecular genetics Powerful techniques for measuring mutational events In addition to causing cancer, physiological
(Sect. XIII) (quantitatively and qualitatively); mutational relevance of somatic mutations (and other
fingerprinting to identify oxidative stress; simple genetic abnormalities) is not obvious
methods applicable to wide array of species
Transgenic organisms Immensely powerful tools; a direct test of free Somewhat expensive and time consuming;
(Sect. XIV) radical theory; interbreeding of transgenic animals dynamic nature of oxidative stress may
for sophisticated in vivo research; established complicate interpretations of phenotypes
methods for widely divergent animals
(Caenorhabditis elegans, Drosophila, mice)
Sporadic degenerative Results (if positive) provide mechanistic insights into Lack of animal models; disease may be of
diseases the biology of oxidative pathophysiology and strictly human relevance (not a general
(Sect. XV) evidence of etiological relevance phenomenon of aging)
Inherited degenerative Results (if positive) provide mechanistic insights into Usually lack of an animal model; phenomenon
diseases the biology of oxidative pathophysiology and (often severe) may not be of relevance either
(Sect. XVI) evidence of etiological relevance to most humans, nor to aging in general
Epidemiology Large data sets permit subtle specieswide trends to Currently expensive, long term, and difficult
(Sect. XVII) be detected; public health and policy relevance
(preventative medicine); may be facilitated by
sequencing of the human genome

cited species of O2 termed singlet oxygens can result site spins, thereby eliminating the spin restriction of
from the absorption of energy (for instance, from ultravio- ground-state O2 and enabling greater reactivity. The chem-
let light). Designated by the formulas 1DgO2 and 1Sg/O2 , istry of oxygen and its derivatives has been extensively
both of these species differ from the triplet ground state discussed elsewhere (115, 342). Because all of these spe-
(3Sg0O2) in having their two unpaired electrons in oppo- cies (O20j, H2O2 , jOH, 1DgO2 , and 1Sg/O2), by different

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 551

routes, are involved in oxygens toxicity, we will collec- from which mSOD has been removed (by sonication of
tively refer to them as oxidants. the intact organelles followed by extensive washing) was
It is now beyond doubt that oxidants are generated used for the detection of ETC O20j. In these experiments,
in vivo and can cause significant harm (20, 37, 60, 91, 112, stoichiometric estimates of the ratio of O20j generation
351). There are numerous sites of oxidant generation, four (by submitochondrial particles) to H2O2 generation (by
of which have attracted much attention: mitochondrial the intact organelles) fell between 1.5 and 2.1 (24, 25, 69,
electron transport, peroxisomal fatty acid metabolism, cy- 89, 191); because two O20j molecules dismutate (either
tochrome P-450 reactions, and phagocytic cells (the res- spontaneously or with the help of mSOD) to form one
piratory burst). Before a discussion of the potential con- molecule of H2O2 , such results suggest that virtually all
tributions of different sources of oxidants, it is worthwhile mitochondrial H2O2 may originate as O20j (27). Moreover,
briefly to outline them. because most cellular H2O2 originates from mitochondria,
In the textbook scheme of mitochondrial respiration, O20j from the ETC may be a cells most significant source
electron transport involves a coordinated four-electron re- of oxidants (37).
duction of O2 to H2O, the electrons being donated by NADH In a recent discussion of the classic in vitro work (88),
or succinate to complexes I and II, respectively, of the mito- some of the original experimenters take issue with the idea
that free O20j exists in mitochondria as a result of normal

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


chondrial electron transport chain (ETC). Ubiquinone (coen-
zyme Q, or UQ), which accepts electrons from complexes flux through the ETC. They point out that in addition to
I (NADH dehydrogenase) and II (succinate dehydrogenase), having removed mSOD from mitochondria, the sonication
undergoes two sequential one-electron reductions to ubi- they employed also resulted in the loss of cytochrome c,
semiquinone and ubiquinol (the Q cycle), ultimately transfer- which rapidly scavenges O20j in vitro and is present in mito-
ring reducing equivalents to the remainder of the electron chondria at local concentrations from 0.5 to 5 mM. In mito-
transport chain: complex III (UQ-cytochrome c reductase), chondria, they argue, mSOD and cytochrome c rapidly
cytochrome c, complex IV (cytochrome-c oxidase), and fi- scavenge O20j (in the matrix and intermembrane spaces,
nally, O2 (115). However, it appears mitochondrial electron respectively). More to the point, the authors stress that
transport is imperfect, and one-electron reduction of O2 to unless the ETC was poisoned with inhibitors such as anti-
form O0 2 j occurs. The spontaneous and enzymatic dismuta- mycin A, O20j generation was not detected in their experi-
tion of O0 2 j yields H2O2 , so a significant by-product of the ments (89, 191). Arguing that mSOD should act to increase
actual sequence of oxidation-reduction reactions may be the the rate of O20j generation in vivo (by accelerating product
generation of O0 2 j and H2O2 . removal by dismutation to H2O2), they suggest that the
How much O20j and H2O2 do mitochondria generate? actual role of mSOD in vivo may be to increase H2O2 gener-
In classic experiments during the 1970s, measurements of ation (with O20j as a rapidly consumed intermediate) (88).
H2O2 generation by isolated mitochondria indicated that it Ultimately, there remains a good deal of uncertainty sur-
is maximal when ADP is limiting and the electron carriers rounding the mechanisms, quantity, and meaning of mito-
are consequently reduced (state 4 respiration) (26). Es- chondrial O20j generation in vivo (228), a mystery which is
timates of state 4 H2O2 generation by pigeon and rat mito- deepened by recent reports of enzymatic nitric oxide (NOj)
chondrial preparations amounted to 12% of total elec- generation in mitochondria (C. Giulivi and C. Richter, per-
tron flow (26, 229). One problem with this estimate of sonal communication). Because O20j and NOj react to form
mitochondrial H2O2 generation is its reliance on the use the oxidant peroxynitrite (ONOO0), mitochondrial O0 2 j gen-
of buffer saturated with air (20% O2). In vivo, the partial eration may soon need to be considered in the light of its
pressure of O2 is 5%, so these calculations may overesti- ability to destroy NOj and form ONOO0, as discussed in
mate the flux of oxidants in vivo. Even disregarding the section XVIIIB.
use of air-saturated buffer, the initial estimate of percent- A second source of oxygen radicals is peroxisomal b-
age ETC flux leading to H2O2 can be challenged on the oxidation of fatty acids, which generates H2O2 as a by-prod-
grounds that in these experiments the concentrations of uct. Peroxisomes possess high concentrations of catalase,
substrates fed to mitochondria were higher than occurs so it is unclear whether or not leakage of H2O2 from peroxi-
physiologically (118, 146). When H2O2 is measured with somes contributes significantly to cytosolic oxidative stress
more physiological concentrations, the flux is 10-fold under normal circumstances. However, a class of nonmuta-
lower (118), and experiments using subcellular fractions genic carcinogens, the peroxisome proliferators, which in-
of SOD-deficient Escherichia coli suggest in vivo leakage crease the number of hepatocellular peroxisomes and result
of 0.1% from the respiratory chain (146). in liver cancer in rodents, also cause oxidative stress and
What proportion of mitochondrial H2O2 ultimately de- damage (7, 157, 177, 230). Interestingly, during the regenera-
rives from ETC O20j generation? Unfortunately, the mea- tion of the liver after partial hepatectomy, there exist peroxi-
surement of O20j generation by intact mitochondria is pre- somes that do not stain for catalase activity (232), hinting
vented by the presence of mitochondrial SOD (mSOD). that during rapid cell proliferation, oxidant leakage from
Therefore, the isolation of submitochondrial particles peroxisomes may be enhanced.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


552 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

Microsomal cytochrome P-450 enzymes metabolize number of intracellular sources of oxidant that have been
xenobiotic compounds, usually of plant origin, by catalyz- identified in a qualitative way, in terms of ranking their
ing their univalent oxidation or reduction. Although these relative importance, the field is in its infancy.
reactions typically involve NADPH and an organic sub-
strate, some of the numerous cytochrome P-450 isozymes
directly reduce O2 to O20j (105, 168) and may cause oxida- C. Targets of Oxidants
tive stress. An alternative route for cytochrome P-450-
mediated oxidation involves redox cycling, in which sub- What are the targets of endogenous oxidants? The
strates accept single electrons from cytochrome P-450 and three main classes of biological macromolecules (lipids,
transfer them to oxygen. This generates O20j and simulta- nucleic acids, and proteins) are susceptible to free radical
neously regenerates the substrate, allowing subsequent attack, and there is plentiful evidence that all suffer oxida-
rounds of O20j generation (115). Although it is unclear to tive damage in vivo. Although it is well beyond the scope
what extent cytochrome P-450 side reactions proceed un- of this review to treat the biochemistry of oxidative dam-
der normal conditions, it is possible that such chronic age in any great depth, the area has been expertly re-
O20j generation by cytochrome P-450 is the price animals viewed (115). A synopsis of the better known pathways

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


pay for their ability to detoxify acute doses of toxins (6). of oxidative damage, however, is warranted; the most fa-
Finally, phagocytic cells attack pathogens with a mix- miliar end products are described here.
ture of oxidants and free radicals, including O20j, H2O2 , The earliest research on the destruction of biological
NOj, and hypochlorite (38, 220, 262). Although the mas- molecules by oxidants involved lipids (109). Food chem-
sive generation of oxidants by immune cells differs from ists have long understood that the rancidity of fats results
the above three sources of free radicals to the extent that from peroxidative chain reactions in lipids (autoxida-
it is the result of pathogenesis, it is nevertheless a normal tion); a lipid hydroperoxyl radical abstracts a hydrogen
and unavoidable consequence of innate immunity. atom from the double bond of a neighboring unsaturated
Chronic inflammation is therefore unique among the en- lipid, forming a hydroperoxide and an alkyl radical, the
dogenous sources of oxidants, because it is mostly pre- latter which combines with O2 to regenerate a lipid hydro-
ventable (49, 231, 245). peroxyl radical capable of initiating another round of oxi-
In addition to these four sources of oxidants, there dation. Ultimately, intramolecular reactions and decom-
exist numerous other enzymes capable of generating oxi- position yield cyclic endoperoxides and unsaturated alde-
dants under normal or pathological conditions, often in hydes, the latter of which are reactive and may act as
a tissue-specific manner (115). To give a single relevant mutagens (194) or inactivate enzymes (39, 322), or operate
example, the deamination of dopamine by monoamine as endogenous fixatives, reacting with proteins and nu-
oxidase generates H2O2 , in some neurons, and has been cleic acids to form heterogeneous cross-links (42). More-
implicated in the etiology of Parkinsons disease (80). over, a primary effect of lipid peroxidation is decreased
Finally, the widespread catalytic generation of NOj, membrane fluidity, which alters membrane properties and
achieved by various isozymes of nitric oxide synthase and can significantly disrupt membrane-bound proteins (324).
central to processes as diverse as vascular regulation, im- Oxidative damage to nucleic acids includes adducts
mune responses, and long-term potentiation, increases of base and sugar groups, single- and double-strand breaks
the potential routes for destructive oxidative reactions in the backbone, and cross-links to other molecules. The
(187). The interaction between O20j and NOj results in spectrum of adducts in mammalian chromatin oxidized in
ONOO0, which is a powerful oxidant. vitro and in vivo includes more than 20 known products,
As originally articulated by Harman (120), the free including damage to all four bases and thymine-tyrosine
radical theory of aging did not distinguish between these cross-links (70, 71, 113). The electrochemical properties
different sources of oxidants. However, the rate of living of the adduct 8-oxo-guanine (oxo8gua) and the deoxy-
hypothesis clearly singled out mitochondrial O20j and H2O2 nucleoside 8-oxo-2,7-dihydro-2*-deoxyguanosine (oxo8dG),
generation, since it is the mitochondrial respiration rate which have permitted the coupling of extremely sensitive
that negatively correlates with MLSP. Also, as many other electrochemical detection to high-performance liquid
established sources of oxidants are tissue specific (associ- chromatography (HPLC), have resulted in hundreds of
ated with hepatic, neuronal, and other specialized func- studies of its formation, accumulation, and excretion (17).
tions), they are less likely to explain aging across a broad The identification of specific enzymatic repair of oxidative
range of species. For this reason, mitochondrial O20j and lesions has recently provided both proof of the signifi-
H2O2 have captured the lions share of attention. However, cance of oxidative DNA damage as well as tools to manip-
it may turn out that for some age-associated disorders, ulate the load of damage in vivo by genetic knockout (17,
nonmitochondrial oxidants are critical. In the expanded 23, 78, 192, 266, 291).
sense of the free radical theory, any oxidants, mitochon- The oxidation of proteins is less well characterized,
drial or not, may play a role. Therefore, despite the great but several classes of damage have been documented,

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 553

including oxidation of sulfhydryl groups, reduction of di- O20j (37). Consequently, the manipulation of O2 partial
sulfides, oxidative adduction of amino acid residues close pressure is a relatively simple tool that has been used to
to metal-binding sites via metal-catalyzed oxidation, reac- test the free radical theory.
tions with aldehydes, protein-protein cross-linking, and
peptide fragmentation (317, 318). A particularly intriguing
recent development has been the realization that a num- D. Antioxidant Defenses
ber of enzymes possessing active-site iron-sulfur clusters
are acutely sensitive to inactivation by O20j (86, 176). For Cells are equipped with an impressive repertoire of
example, E. coli aconitase is inactivated by O20j with a antioxidant enzymes, as well as small antioxidant mole-
rate constant of 109 M01rs01 (95, 96). Mammalian mito- cules mostly derived from dietary fruits and vegetables
chondrial aconitase is inactivated in vitro and in vivo by (5, 351). These include 1) enzymatic scavengers such as
treatments that increase mitochondrial O20j generation, SOD, which hastens the dismutation of O20j to H2O2 , and
such as growth under hyperbaric conditions (97, 98). Be- catalase and glutathione peroxidase (GPX), which convert
cause aconitase participates in the citric acid cycle, its H2O2 to water; 2) hydrophilic radical scavengers such as
inhibition would be expected to have pleiotropic effects. ascorbate, urate, and glutathione (GSH); 3) lipophilic radi-
Moreover, the mechanism of aconitase inhibition by O20j

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


cal scavengers such as tocopherols, flavonoids, carot-
has been demonstrated to involve the release of free iron enoids, and ubiquinol; 4) enzymes involved in the reduc-
from the enzyme (86). Free iron atoms catalytically exac- tion of oxidized forms of small molecular antioxidants
erbate oxygen stress (see below), and it has been pro- (GSH reductase, dehydroascorbate reductase) or respon-
posed that superoxides genotoxicity is a function of its sible for the maintenance of protein thiols (thioredoxin
ability to liberate protein-bound iron (159, 184). reductase); and 5) the cellular machinery that maintains
Unlike lipids and nucleic acids, proteins represent a a reducing environment (e.g., glucose-6-phosphate dehy-
very diverse target for oxidative damage. Although protein drogenase, which regenerates NADPH). The complement
oxidation has been demonstrated at the level of the pep- of defenses deployed differs not only between organisms
tide backbone and amino acids, there has been relatively or tissues, but even between cellular compartments. For
little scrutiny of differences between proteins in their sen- instance, GPX plays an important role in mammals but is
sitivities. A detailed quantitative comparison of bovine absent from flies and nematodes (298, 330), and there
serum albumin and glutamine synthase has shown suscep- exist in humans three forms of SOD (cytosolic Cu,Zn-SOD,
tible residues of the former (methionine and the aromatic mitochondrial Mn-SOD, and extracellular SOD), encoded
amino acid residues) to be oxidized about twice as fast and regulated independently (91).
as those on the latter, implicating all four levels of protein As far as supporting the free radical theory of aging
structure in relative susceptibility (19). A study of the is concerned, the universality of antioxidant defenses is
oxidation sensitivities of a various cloned K/ channels good news. Although the nature of these defenses varies
from T lymphocytes, cardiac cells, and neurons revealed between species, the presence of some type of antioxidant
that whereas five of the cloned channels were highly sensi- defense is universal. In fact, some antioxidants, such as
tive to oxidation, an equal number were resistant (74). SOD, are very highly conserved. Clearly, an indifference
Differential sensitivities raise the possibility that the loss to oxygen free radicals is inconsistent with life, underlin-
of homeostasis that is a hallmark of aging could result ing the centrality of oxidative damage. Moreover, the fact
from the selective oxidation of proteins. that antioxidant defenses are not uniform has been incor-
In the context of aging, a particularly relevant aspect porated into the free radical theory; differences in antioxi-
of oxygens toxicity is its promotion by some metals and dant defenses between species have been put forth to
by elevated O2 partial pressure. Iron and copper catalyze explain differences in life span. Although there is some-
the homolytic cleavage of ROOH (the Fenton reaction), thing uncomfortably ad hoc in these two different inter-
leading to the generation of jOH (115). It is jOH that is pretations of the data, they are not inconsistent. Whereas
the most reactive oxidant, reacting at diffusion-limited aerobic life requires organisms to cope with oxidation to
rates. The catalytic properties of iron and copper explain some extent, different evolutionary pressures appear to
why cells possess metal-chelating proteins such as ferritin have selected for more or less investment in these de-
and transferrin, which reduce the concentration of redox- fenses, as is discussed in section III.
active metals (114, 211). In humans, the bodys content Finally, a persistent problem in testing the free radi-
of iron increases with age (in men throughout their lives, cal theory is that antioxidants are both parallel (different
and in women after menopause), and it has been sug- antioxidants can play similar roles, e.g., catalase and GPX)
gested that this accumulation may increase the risk of and serial (enzymes operate in tandem to decompose radi-
oxidative damage with age (169, 331). Finally, oxidative cals to harmless products, e.g., SOD and catalase). Conse-
stress in vivo is aggravated by increasing O2 partial pres- quently, measurements of individual antioxidant activities
sure, due to a more pronounced flux of mitochondrial do not have great relevance. In fact, as is discussed below,

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


554 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

measurements of age-related changes in individual antiox- sponse to oxidative challenges (67, 119, 320) and are of
idants have led to conflicting results (297). For this reason, course potential targets of oxidative destruction (145).
aggregate assays have been devised, such as the suscepti- Also, the generation of oxidants may be enhanced by the
bility of crude cellular homogenates to in vitro oxidation malfunctioning of oxidatively damaged molecules (28,
by ionizing radiation (2, 300). Although these assays do 303). Therefore, with the examination of Figure 1, it is
not provide any information about the specific mecha- not difficult to envision ways in which primary oxidative
nisms of defense, they conveniently measure overall effec- destruction of any target (e.g., the components of the
tiveness. mitochondrial ETC, scavenging enzymes such as SOD, or
DNA repair enzymes) might promote further oxidative
damage in what is frequently called a catastrophic vi-
E. Repair of Oxidative Damage
cious cycle.
Although such cycles are intuitively appealing, their
Unlike defenses against oxidants, which have been
documentation awaits future work and will be extremely
extensively characterized, the machinery for repairing ox-
difficult from a technical standpoint. An alternative to lab-
idative damage is relatively unexplored. Nevertheless, it
based approaches, namely, the computational modeling
is clear that cells repair oxidized lipids (e.g., phospholi-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


of these complex interactions in what has been termed a
pase A2 cleaves lipid peroxides from phospholipids; Ref.
Network Theory of Ageing, is being pursued by theoreti-
239), oxidized nucleic acids (e.g., glycosylases specifically
cal gerontologists (170). Ultimately, a question of obvious
recognize and excise oxidized bases from double-
importance is whether or not such cycles, if they exist,
stranded DNA; Refs. 23, 56a, 56b, 325), and oxidized pro-
could be broken, and modeling may help pinpoint weak
teins (104, 240, 261, 316). The comparative biochemistry
links for therapeutic intervention.
of cellular repair is very fertile ground for the free radical
theory.
III. REFINEMENTS AND COROLLARIES OF
THE FREE RADICAL THEORY
F. Synthesis: Interaction of Oxidant Generation,
Oxidative Damage, and Repair
As the free radical theory has gained ground, it has
incorporated other ideas. For example, as mentioned
The existence of multiple intracellular sources of oxi-
above, the rate of living hypothesis dovetailed with the
dants and complex defenses has led to refinements of
free radical theory once mitochondrial free radical genera-
the free radical theory. For example, it is clear that the
tion was confirmed. Three other ideas that have been
metabolism of oxygen radicals is dynamic, with damage
influential are the evolutionary concept of antagonistic
resulting from an increase in oxidant generation or a de-
pleiotropy, the somatic mutation theory of aging, and the
crease in antioxidant defenses. Consequently, a difference
mitochondrial theory of aging.
in life span between species or individuals could be due
to different rates of living, or to different rates of scav-
enging (57, 82). The picture has been further complicated A. Oxidants and Evolutionary Theories of Aging
by the discovery of specific enzymatic repair of oxidative
damage, leading to repair or fidelity versions of the The intracellular generation of oxidants capable of
theory, in which life span is determined by the failure to limiting life span may appear paradoxical. It seems rea-
correct oxidative damage (137, 239). sonable to expect that natural selection might have de-
The relationship among these three components of vised aerobic cells that do not leak toxic by-products.
oxidative stress: oxidant generation, antioxidant protec- Evolutionary biologists have contributed to the free radi-
tion, and repair of oxidative damage, and the way in which cal theory by suggesting why physiologically harmful gen-
they have been investigated in testing the free radical eration of oxygen radicals occurs. They have argued that
theory, is illustrated schematically in Figure 1. Increases natural selection favors genes that act early in life and
in oxidant generation, and decreases in antioxidant pro- increase reproduction, rather than genes that act to pre-
tection and repair systems, are among the theorys testa- serve nongerm cells (the disposable soma), a principle
ble predictions and have been examined both as a func- called antagonistic pleiotropy (162164, 265, 341).
tion of age in individuals of the same species, as well as The concept of antagonistic pleiotropy stresses that
between species of differing MLSP. in the wild, reproductive success is principally a function
Finally, an extremely important (if experimentally re- of external factors. With the exception of modern-day
calcitrant) aspect of the interactions between oxidants, humans, individuals do not usually die of old age, but are
antioxidants, and repair are feedback loops, positive and eaten, parasitized, or out-competed by others. Therefore,
negative, between them. Antioxidant defenses and cellu- preserving the cells of the disposable soma, otherwise
lar repair systems have been shown to be induced in re- known as the body, may be disadvantageous if it detracts

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 555

FIG. 1. The ultimate outcome of oxidative


stress is a function of 1) oxidant generation, 2)
antioxidant defenses, and 3) repair of oxidative
damage. Bolds arrows denote oxidative damage,
and dashed arrows denote routes for its preven-
tion or repair. Because of the ways in which
these processes may interact, multiple positive-
and negative-feedback loops are possible. Aging
(A) is situated at intersection of these processes.
In testing the free radical theory, changes in pro-
cesses 13 have been measured both as a func-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


tion of age and as a function of species maxi-
mum life span potential. The similarity of the
figure to the international emblem of radiation
is not a coincidence; the free radical theory has
its roots in radiation biology.

from more pressing problems. The toxicity of respiratory remains to be seen whether or not the argument is valid
burst oxidants, for instance, cannot easily be eliminated for nonproliferative senescence. For instance, whereas
by evolution, since this would result in death from child- significant age-related increases in somatic mutations in a
hood infection. Similarly, an investment in improved anti- reporter transgene (lacZ) have been measured in a mitotic
oxidant defenses maximizes fitness only if the resources tissue of transgenic mice (the liver), no increase was de-
are not better invested in strength, beauty, speed, or cun- tected in the largely postmitotic brain of the same animals
ning. (71a), suggesting that neurodegeneration, at least, is un-
In terms of natural selection, the tremendous cost of likely to be the result of accumulated somatic mutations
death before reproductive age, the constantly compound- in nuclear DNA. Moreover, the accumulation of mutations
ing probability of death from external threats, and the cost in the liver tissues was not dramatic, suggesting that muta-
of failing to reproduce all ensure that selective pressure is genesis may be of little functional consequence to mitotic
strongest at young ages. Any novel mutations that de- tissues as well (338b). In light of these data, what evidence
crease oxidative damage first have to satisfy the criteria is there that somatic mutations are related to aging? A
of youthful reproduction. In short, selective pressure to compelling argument for the somatic mutation theory of
compete effectively at an early age may guarantee a cer- aging was provided years ago in the discovery that DNA
tain degree of O2 toxicity and work against the conserva- repair ability correlates with species-specific life span
tion of the soma in the long run. (127a), a phenomenon that has recently been reconfirmed
(52a). However, it has been noted that DNA repair, which
is necessary for the prevention of tumorigenesis, is neces-
B. Oxidants and the Somatic Mutation Theory sary but not sufficient for longevity (52a). Ultimately, ar-
of Aging guments about the physiological significance of somatic
mutations hinge on how disruptive a given mutational
The somatic mutation theory holds that the accumu- burden is to a cell or animal; with current methods, this
lation of DNA mutations is responsible for degenerative is an unanswerable question.
senescence (23, 79, 212, 218, 332). In the case of cancer, In any case, it has been demonstrated in numerous
which results from both point mutations in oncogenes studies with prokaryotes, yeast, and mammalian cells that
and the loss of tumor suppressor gene function (often by oxidants are mutagens, against which cells actively pro-
deletion), the role of mutations is unquestionable (5). It tect their genetic material (81, 108). Although it is not

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


556 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

yet clear what fraction of mutations can be attributed oxidative adducts in DNA), and depletion (such as the
to oxidative damage, the characterization and cloning of loss of enzymatic activity or reduced thiols).
defense genes against oxidative mutagenesis (17), and the
development of in vivo mutagenesis assays (198), has fi-
nally opened up avenues for definitive experiments. A. Accumulation of Oxidative End Products

The gradual and steady accumulation of intracellular


C. Oxidants and Mitochondrial Theories of Aging
yellow-brown fluorescent pigments, referred to as lipofus-
cin, occurs in numerous phyla. Lipofuscin arises promi-
The mitochondrion has also long attracted attention
nently in postmitotic cells (where, it is argued, it remains
as one of the cells weak links, an organelle whose dys-
undiluted by rounds of cell division; Ref. 296) and is lo-
function has profound negative pleiotropic effects (193).
cated in small granules in secondary lysosomes. Lipofus-
Mitochondria supply ATP and also sequester potentially
cin is structurally complex and variable, consisting mostly
toxic Ca2/, yet because of their generation of O20j and
of cross-linked lipid and protein residues (251, 296, 327),
H2O2 , they are on the front lines of respiratory oxidative
and is ubiquitous, documented in species as diverse as
stress. The idea that the mitochondrion is therefore

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


nematodes, fruit flies, rats, bees, crab-eating monkeys, and
uniquely vulnerable was embraced early on by proponents
crayfish. Most important, it is abundant in aged tissues,
of the free radical theory (121). In the early 1980s, Miquel
where it may occupy more than one-half of the volume
and colleagues (84, 212, 213) proposed that oxidative dam-
of the cell (347, 348).
age to mitochondrial DNA (mtDNA) in postmitotic cells
Early on, it was discovered that incubation of amino
would lead to mutations and blocks to replication, and
acids with the lipid peroxidation product malonaldehyde
consequently to mitochondrial dysfunction and physiolog-
under acidic conditions leads to the formation of lipofus-
ical decline. This mtDNA mutation hypothesis of aging,
cin-like fluorophores (42). The plausibility of such a reac-
which incorporates free radicals, somatic mutations, and
tion, given the contents and low pH of lysosomes, sug-
the central role of mitochondria in homeostasis, is pres-
gested that lipid peroxidation in vivo leads to the forma-
ently under intense scrutiny (8, 11, 21, 50, 110, 222, 223,
tion of lipofuscin (324). Other in vitro studies of lipid
238, 258260, 276, 288, 336, 339).
peroxidation have since uncovered a great number of
routes to fluorescent, cross-linked products via promiscu-
IV. OXIDATIVE PHENOMENOLOGY: ous oxidative chemistry, which suggests that lipofuscin is
AGE-ASSOCIATED TRENDS a biomarker of lipid peroxidation (160, 296, 348).
Despite extensive in vitro experiments, it is not
The phenomenological approach to the free radical known with certainty how lipofuscinogenesis occurs in
theory has involved looking for traces of oxidative dam- vivo, nor how lipofuscin comes to accumulate with age.
age in vivo. Phenomenology is not well suited to critically Lipid peroxidation could occur throughout the cell and
testing the free radical theory, since the data (which are be followed by lysosomal phagocytosis and cross-linking
voluminous and generally supportive) mainly represent of peroxidative by-products, in which case an age-related
correlations. Documented increases in oxidative damage, increase in lipofuscin content could be seen as the result
no matter how impressive, may be a consequence of a of oxidative damage. Alternatively, an age-associated de-
primary nonoxidative event. Nevertheless, phenomenol- cline in lysosomal activity (due to something besides oxi-
ogy is the foundation upon which more powerful experi- dation) might increase the residence time of phagocy-
ments depend, since the analytical methods developed for tosed material enough to enhance lipofuscinogenesis in
it have been used to compare species, genetic mutants, situ from constant amounts of peroxides (135). In support
and populations with differing life spans. In fact, almost of the latter possibility, infusion into rat brains of lyso-
all of the biomarkers of oxidative stress described in this somal proteinase inhibitors leads to the rapid accumula-
section have been found to accumulate at a faster rate in tion of lipofuscin-like granules (149). This scenario sug-
short-lived species, and in many cases, this rate correlates gests that lipofuscinogenesis may be a consequence, not
with O2 consumption. Familiarity with the most frequently a cause, of aging.
measured end points is a prerequisite to assessing the free Experiments with cultured cardiac myocytes have es-
radical theory. tablished roles for both oxidative damage and lysosomal
If oxidative damage is a significant cause of cellular turnover (28). Lipofuscin accumulates in these cells in
degeneration, then one expects to see more of it in older culture, and growth under increasing O2 partial pressure
individuals. Oxidative damage has been described in from 5 to 40% markedly enhances its accumulation (306).
terms of accumulation, modification, and depletion: ac- Inclusion of iron in the growth medium further increases
cumulation of end products of oxidative damage (such as lipofuscinogenesis, and the iron chelator desferal de-
lipofuscin), modification of existing structures (such as presses it, suggesting that Fenton reaction-generated jOH

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 557

is an initiator (200). Finally, antioxidants inhibit lipofuscin formation of which requires the presence of O2 (326). It
formation in cultured cardiomyocytes, whereas lysosomal appears as if Amadori products themselves are a source
protease inhibitors increase it (199, 201, 202). of H2O2 in vitro, which then accelerates glucose-mediated
Even if oxidative damage is primarily responsible for fluorogenic collagen cross-linking in a catalase-sensitive
depositing lipofuscin in the lysosomes of senescing ani- fashion, although it is unclear to what extent this occurs
mals, is it more than a biomarker of aging? It has been in vivo (77, 217). As is the case with other AGEs, the
theorized that lipofuscin accumulation is likely to impair tissue burden of pentosidine is elevated in diabetics, as a
autophagy, as more lysosomal volume is occupied by the consequence of hyperglycemia.
indigestible material (28). Because lysosomes are respon- Pentosidine has been found to accumulate as a func-
sible for the recycling of materials and organelles, their tion of age in shrews, rats, dogs, cows, pigs, monkeys,
failure may include the following: 1) a delay in mitochon- and humans, yielding equivalently shaped curves in all
drial turnover (with a concomitant decrease in mitochon- cases (284). It is not clear, however, how glycooxidative
drial efficiency or an increase in mitochondrial oxidant modifications might contribute to degeneration. It has
generation), 2) an accumulation of oxidatively modified been proposed that cross-linking in cartilage is related to
proteins and lipids in the cytosol awaiting degradation its decreased elasticity and relative resistance to proteoly-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


(potentially aggravating cytosolic lipid peroxidation), 3) sis in old animals (9). However, the absolute amount of
an accumulation of lipofuscin-bound iron in a redox-ac- collagen pentosidine cross-links attained at death is much
tive form (which might promote further intralysosomal higher in long-lived than in short-lived species: 67 pmol/
lipid peroxidation), and 4) the disruption of lysosomal mg in 3.5-yr-old shrews, 1518 pmol/mg in 25-yr-old mon-
membranes (and the spillage of hydrolytic enzymes into keys, and 50100 pmol/mg in 90-yr-old humans (284). In
the cytosol). Although these speculations (28) remain to other words, it appears that the rate of pentosidine accu-
be substantiated, it has been shown that when treated mulation may merely be a measure of more rapid oxida-
with sublethal doses of H2O2 , cultured cells display lyso- tive damage in short-lived species rather than an actual
somal disruption and leakage of the lysosomal compart- cause of dysfunction.
ment into the cytosol (29). Also, it has been demonstrated An intriguing twist to this story has been the cloning
that the sensitivity of cultured primary hepatocytes to of a specific cellular receptor for AGE, called RAGE (re-
oxidation, which was associated with a loss of GSH and an ceptor for AGE), that belongs to the immunoglobulin su-
influx of Ca2/, was prevented by the iron chelator desferal perfamily and is expressed by mononuclear cells and the
(235). What is intriguing about these results is the fact vascular endothelium (277, 278). One of the effects of
that whereas desferal stabilized the lysosomes, it did not AGE binding by RAGE is the generation (in mononuclear
prevent the loss of GSH or the increase in intracellular cells) of intracellular oxidants, the activation of the oxi-
calcium, so it may be that it is lysosomal leakage per se, dant-sensitive transcription factor NFkB, and the induc-
rather than peroxidative damage, which is the actual le- tion of downstream events linked to atherogenesis (279,
thal step in this model of oxidative killing. 280). As discussed in section XVB, it has recently been
shown that RAGE, which is highly expressed by microglial
cells in the brain (142), is a receptor for amyloid b-peptide
B. Steady-State Levels of Oxidative Modification (Ab). The RAGE binding of Ab results in oxidant genera-
tion, implicated in the etiology of AD (293, 344).
Unlike cytosolic proteins whose half-lives are mea- Several amino acid residues in proteins are suscepti-
sured in minutes or hours, some extracellular proteins are ble to oxidative modification, forming side chain carbonyl
rarely recycled, and oxidative modification of these old derivatives (317). The development of sensitive methods
macromolecules occurs. A class of fluorescent cross- for the analysis of protein carbonyls by Stadtman and co-
linked molecules that is distinct from lipofuscin forms on workers (181) enabled them to study oxidative modifica-
long-lived proteins such as collagen and lens crystallin tion in human brain tissue and cultured fibroblasts, and
(216). These modifications are initiated by the reaction in rat liver. They found a two- to threefold rise in protein
of reducing sugars with free amino groups (glycation), a carbonyl content between young and old age, an increase
chemical sequence that is unrelated to oxidation and re- from 10 to 30% of the total protein pool (315). The in-
sults in a molecule known as an Amadori product. Further crease was exponential and correlated well with de-
nonoxidative rearrangements result in stable, cross-linked creased activity of the oxidation-sensitive enzyme glu-
advanced glycation end products (AGEs) (35), whose ab- cose-6-phosphate dehydrogenase (G-6-PD). In compari-
solute abundance appears to be an excellent biomarker son, the rise in protein oxidation in the mongolian gerbil
of age (217). Recently, it was discovered that oxidation was less dramatic, increasingly significantly in brain,
is one fate of the Amadori product. Pentosidine, the name heart, and testis, but not in kidney. As in human tissues,
given to a cross-link involving arginine, lysine, and pen- trends for the activity of G-6-PD correspond to the in-
tose moieties, is one such glycoxidation product, the creased damage, falling in brain and heart but not in kid-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


558 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

ney (300). Similar results have been reported in an insect increase its degradation by proteases (322). This differ-
model. An age-associated 2.5-fold increase in the protein ence exists despite the fact that in both cases, the same
carbonyl content of old versus young houseflies has also lysine residue is affected. To make matters more complex,
been documented (299), and as in humans, the increase the cross-linking of G-6-PD multimers by 4-hydroxy-2-non-
occurs exponentially during the life span. The similarity enal (which predictably results in a product with lipofus-
of the degree and pattern of increase in insects and mam- cin-like fluorescence) produces a molecular species that
mals is striking, considering the enormous difference in actually inhibits the multicatalytic protease (92). The
their MLSP (40 days vs. 100 yr). Moreover, protein car- physiological cost of protein oxidation is presently an un-
bonyl levels increase similarly in mitochondrial extracts known quantity.
from the thoracic flight muscles of these animals (303). The appearance of protein-bound 3,4-dihydroxyphe-
Mitochondrial aconitase is particularly prone to oxidative nylalanine (DOPA) on jOH-damaged proteins has been
modification during aging in vivo and was identified by characterized; when converted to a quinone, protein-
the immunoblotting of housefly mitochondrial protein ex- bound DOPA can undergo redox cycling, generating
tracts with a monoclonal antibody designed to detect pro- O20j. It has therefore been proposed that protein oxidation
tein carbonyls (343b). Carbonylation of this key citric acid may contribute to the progression of aging not merely by

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


cycle enzyme increased in parallel with a decline in its the loss of protein function, but also by an acceleration
activity. of the flux of oxidants (61, 63, 64, 69, 101, 102).
Somewhat stronger evidence that protein oxidation The oxidative modification of DNA has also been
may play a causative role in senescence comes from com- studied in animals of different ages, with conflicting re-
parisons of crawlers versus fliers of the same aging sults. Although some studies have reported a modest in-
cohort. Although the two groups share the same chrono- crease in specific oxidative adducts, single-strand breaks,
logical age, crawlers are phenotypically senescent individ- and abasic sites, others have been negative (23, 132, 156,
uals that have lost the ability to fly and have a shorter 221, 337). The failure to detect an age-related increase
remaining average life span than do fliers (e.g., 9.0 days in oxidative adducts by the analytical chromatographic
vs. 13.3 days for 10-day-old crawlers and fliers, respec- techniques typically employed may have been due to the
tively). The protein carbonyl content of crawlers was 29% difficulty of working close to the limit of sensitivity (17).
higher than that of fliers (299), reflecting their greater In fact, it has become apparent that the measurement of
phenotypic age, as was the degree of carbonyl modifica- the adduct oxo8dG is frequently plagued by artifacts (29a,
tion of mitochondrial (but not cytosolic) aconitase (343b). 44b, 127c, 156a, 248a) and that these may have compro-
Humans suffering from Werners syndrome, a disease mised some published experiments. Of particular concern
characterized by premature senescence, are individuals are measurements of oxo8dG in mtDNA (16), which have
whose phenotypic aging is also accelerated, and they too generally been higher than in nuclear DNA, but which
appear to have more extensive protein oxidation. Fibro- may be particularly prone to artifacts associated with the
blasts from Werners patients of all ages have a level of analysis of small samples (16, 127c). Moreover, it is note-
protein carbonyls equivalent to that in 80-yr-old controls worthy that even among the highly variable published esti-
(233). In a creative study attempting to correlate protein mates of oxo8dG in mtDNA are values that are equivalent
oxidation to a physiologically relevant end point, it was to the lowest measured values of oxo8dG in nuclear DNA
shown that in old mice, interanimal variation in protein (131a). Because of the small number of studies of mtDNA
carbonyl content of two different areas of the brain (cere- and the high variability between the measured values, it
bral cortex vs. cerebellum) was associated with parallel is not yet possible to conclude that mtDNA is, in fact,
interanimal variation in memory and motor function defi- more heavily oxidized than nDNA. Encouragingly, alterna-
cits (90). tive PCR-based methods for measuring oxidative damage
Are protein carbonyls physiologically relevant, or are have recently been used to compare oxidation of mtDNA
they merely markers? What are the actual consequences and nDNA by exogenous oxidants, with the result that
of protein modification? Unfortunately, there are few the former appears more sensitive than the latter (270a,
quantitative data with which to answer this question, al- 343a), although these studies could not quantify baseline
though qualitative data exist. The fate of oxidized proteins values of damage. With methodological improvements, fu-
may depend on the form of damage. For example, metal- ture experiments may be more conclusive. For instance,
catalyzed oxidation of G-6-PD by iron/citrate results in a the use of single-cell gel electrophoresis (the comet assay)
thermolabile enzyme that is a better substrate for proteol- to measure single-strand breaks and abasic sites in whole
ysis than is the native enzyme (93). Rapid turnover of rat hepatocytes in situ revealed a statistically significant
metal-oxidized G-6-PD may therefore proceed efficiently. 1.5-fold increase in old rats compared with young rats
On the other hand, G-6-PD modification by 4-hydroxy-2- (131) (although this experiment did not distinguish be-
nonenal, a lipid peroxidation product, also inactivates the tween oxidative and nonoxidative damage).
enzyme but does not render the enzyme thermolabile or In any case, even if the burden of oxidative adducts

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 559

does increase with age, there is virtually no information repair, but studies that have measured age-related
about the likely effect of oxidative DNA damage in vivo, changes in antioxidant defenses have generated conflict-
apart from the knowledge that it leads to mutations and ing results. Recent measurements of antioxidants in mon-
cancer. The fact that there is active DNA repair in postmi- golian gerbils (300) and mice (215) are representative of
totic tissues (in which the danger of mutation due to repli- the types of patterns that have been uncovered in many
cation is nonexistent), and that such repair is often tar- other studies (65, 248, 254, 263, 274, 301, 302, 305, 329).
geted to transcribed regions of the genome, suggests that In various tissues of gerbils, there was not a consistent
DNA damage itself interferes with gene expression and is pattern of change; increases in SOD and decreases in GSH
not tolerated (116, 117). This important question deserves were observed, whereas GPX was equivalent at different
more attention. ages and catalase increased or decreased, depending on
the tissues and the age at analysis. In mouse brain, on the
C. Oxidative Depletion of Biochemical Pools other hand, significant decreases in SOD, catalase, and
GSH reductase were observed, although GPX levels were
The oxidative depletion of molecules with increasing unchanged.
age has not been well documented in senescent animals, Another complication is that defenses are induced in

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


since the destruction of molecules does not often leave response to stress. Therefore, a higher level may indicate
traces; luckily, some pathways of oxidative damage do better protection, or alternatively, greater need for antiox-
leave biochemical fingerprints. The loss of integrity of idant defenses due to an increase in oxidant generation.
lipid bilayers due to peroxidation is one of the most salient Studies of antioxidants in rat heart and skeletal muscles
effects of oxidative damage (324) and results in the gener- illustrate this point. In heart, decreases in cytosolic SOD
ation of aldehydes and alkanes. Unfortunately, these are and GPX and increases in mitochondrial SOD and GPX
not easily measured, the widespread use of the simple were noted in older animals, and several indexes of oxida-
and nonspecific thiobarbituric acid test notwithstanding tive damage were also elevated (151). From these results,
(109). Nevertheless, countless studies have reported an it was concluded that although overall myocardial antioxi-
increase in thiobarbituric acid-reactive substances dant defenses were weakened in the older animals, they
(TBARS) with age. Combined with other more reliable were induced in mitochondria as a compensatory re-
assays, these studies have demonstrated that there is a sponse. In skeletal muscles, in contrast, increases were
greater degree of lipid peroxidation in older animals (209). observed in both cytosolic and mitochondrial forms of all
The measurements of exhaled ethane and pentane is a of the enzymes studied (150), despite the fact that indexes
technique that has the advantage of being applicable to of lipid peroxidation were again elevated; in this case,
humans (165). Unlike lipofuscin and TBARS, which mea- it was concluded that both cytosolic and mitochondrial
sure the size of a pool of destroyed molecules and require antioxidants were induced. The credibility of these
a tissue biopsy, the assay of exhaled hydrocarbons mea- hypotheses is not in question, but it is hard to see how
sures the rate of damage and is noninvasive. Breath pen- they could be disproved. When these and similar studies
tane has been found to increase significantly with age in of age-related antioxidant levels are combined, what re-
humans, suggesting that increased lipid turnover occurs mains is a confusing assemblage of ambiguous trends.
with age because of peroxidation (161a, 208, 356). Re- Of course, interactions between antioxidants are
finement of the technique and elimination of the associ- complex, which aggravates the problem. To avoid the
ated artifacts (314) should facilitate further testing of the problems posed by assays of individual antioxidants, ag-
free radical theory in humans. gregate measures of antioxidant defenses have been de-
The loss of activity of several oxygen-sensitive en- vised. A crude but integrative measure of antioxidant de-
zymes (G-6-PD, glutamate synthetase) has been reported fenses, for instance, is the susceptibility of a homogenate
in mammalian models of aging (315). In houseflies, a de- to induced oxidation. X-irradiation of a whole body ho-
cline in G-6-PD, glutamate synthetase, and alcohol dehy- mogenate of houseflies results in a linear, dose-dependent
drogenase activities has also been documented and coin- increase in protein carbonyls. When homogenates of old
cides with a dramatic loss of protein sulfhydryls (3). An- and young flies are compared, the rate of induction of
other commonly reported age-related loss is an increase protein carbonyls by X-irradiation is 45% higher in 14-
in the ratio of oxidized to reduced glutathione, which may than 5-day-old flies. This suggests that the antioxidant
reflect a disruption of the cells redox state (215, 273). defenses in older flies are less able to cope with oxidative
stress. Moreover, the activity of G-6-PD, an enzyme known
D. Age-Associated Trends in Antioxidant Defenses to be sensitive to oxidation, decreases upon X-irradiation
and Repair of living flies, and does so to a greater extent in old than
young animals (2). When this assay was applied to the
What is the cause of age-related oxidative damage? gerbil samples described above, in which no overall
It could result from less active antioxidant defenses and change in antioxidants was seen, a clear difference be-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


560 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

tween young and old tissues emerged. Whereas 6 krad of ing decrease in protein carbonyls to initial levels. In old
X-irradiation induced a 2038% increase in protein car- animals, on the other hand, no increase in activity was
bonyls in 5-mo-old animals, it induced a 152211% in- observed, and protein carbonyl levels continued to rise
crease in 26-mo-old animals (300). Similarly, although syn- throughout the time course (318).
aptosomes from young and old mice contain equivalent There is circumstantial evidence from mutagenesis
amounts of ATP and GSH, those of old mice were far studies that either antioxidant defenses or repair of oxida-
more sensitive to GSH depletion by the diethyl maleate tive DNA damage (or both) is less efficient in old mice.
than those from young mice (197). Lastly, reperfusion in- The induction of somatic mutations in mice by g-irradia-
jury is a well-established model of oxidative stress associ- tion is from 2.3- to 3.6-fold higher in old than in young
ated with the reestablishment of blood flow following animals, depending on the dose (99). The induction of
ischemia, and it causes greater oxidative damage to heart mutations in young and old animals was reduced by feed-
tissues of old rats than young ones (192a). The use of a ing the animals a cocktail of dietary antioxidants, con-
polyclonal antiserum specific for adducts between lipid firming that oxidants played a mutagenic role in these
peroxidation end products and proteins detected such co- experiments. Therefore, the more pronounced induction
valent modifications of mitochondrial proteins from old of mutations in older mice is indirect evidence of de-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


but not young animals, which was associated with a more creased antioxidant defenses and repair (99). Later exper-
dramatic loss of respiratory capacity in the former. iments employing peripheral lymphocytes from young and
Whereas the baseline mitochondrial respiratory parame- old human subjects resulted in similar results (100). The
ters (before ischemia-reperfusion) did not differ between ability of human peripheral lymphocytes to repair oxida-
young and old animals, the administration of a physiologi- tive DNA damage induced by H2O2 has also been found
cally relevant stress revealed a probable age-related de- to be less efficient in cells from older donors (14).
cline in antioxidant defenses. Altogether, the results above suggest that older cells
Another alternative to measuring absolute levels of may be less able to prevent oxidative damage from oc-
antioxidants in old versus young animals is to investigate curring, and less effective at removing the damage once
the ability of animals of different ages to induce antioxi- it has occurred. There is a clear need for more and better
dants, an approach that has been applied to the analysis data about age-related trends in defenses and repair.
of SOD in the nematode Caenorhabditis elegans (59).
Whereas in young animals challenge with hyperoxia or
the redox cycling compound plumbagin resulted in an E. Age-Associated Trends in Oxidant Generation
increase in SOD activity, in middle-aged or old animals it
actually resulted in a net loss of activity. The accumulation of oxidative damage could also re-
What about repair of oxidative damage? Does its ac- sult from an age-associated increase in the primary gener-
tivity decrease with age? The bulk of evidence suggests ation of oxidants, and some research suggests that this is
that there is probably not an overall age-associated the case. Generation of H2O2 and O2j0 by isolated mito-
change in the intrinsic ability of cells to degrade damaged chondria and submitochondrial particles from 25-mo-old
proteins (94, 270). Although a dramatic decrease in the gerbils, for instance, is 150200% that of 5-mo-old ani-
activity of the oxidized protein-specific alkaline protease mals (300), and that of aged rat heart (305) and brain
has been reported in old rat hepatocytes (318), no change (93a) has also been reported to be elevated. On the other
in this activity was measured in the heart or brain of 25- hand, a recent study that paid specific attention to main-
versus 5-mo-old gerbils (300). In a separate work, a 50% taining physiological substrate concentrations during in
age-related decline in a single activity (peptidylglutamyl- vitro mitochondrial incubations failed to detect an age-
peptide hydrolase activity) of the hepatic multicatalytic related increase in mitochondria H2O2 output (118). Re-
protease was associated with its selective sensitivity (rela- cent measurements of oxidant generation in carefully
tive to the multicatalytic proteases other activities) to isolated rat hepatocytes from young and old animals, em-
metal-catalyzed oxidation, suggesting that resistance to ploying an intracellular dye that fluoresces upon oxida-
oxidants may (logically) characterize the proteases re- tion, have confirmed under conditions that preserve cellu-
sponsible for degrading damaged proteins (46). lar integrity that cellular oxidant generation appears to
Although the intrinsic protease activity may not de- increase (110). A similar increase in oxidant generation
crease with age, there is evidence that repair of oxidized documented in flight muscle mitochondria of houseflies
proteins may be less easily induced in response to an was associated with increases in the activities of every
oxidative insult in old animals. For instance, exposure of measured component of the electron transport chain ex-
young and old rats to 100% O2 increased the content of cept for the content of UQ (308), suggesting that an imbal-
protein carbonyls in both groups over a 48-h period. Be- ance in electron transport may be the cause of aberrant
tween 48 and 54 h of exposure, however, alkaline protease reduction of oxygen. In another experiment, a population
activity was induced in young animals, with a correspond- of chronologically identical 12-day-old flies was separated

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 561

into phenotypically older crawlers and younger fliers as The excretion of the oxo8gua and oxo8dG in urine
described above; mitochondrial H2O2 generation was is a reflection of whole body oxidative hits to DNA (287).
twice as high in the crawlers than in the fliers, reflecting The validity of its use to measure in vivo oxidative hits is
their greater phenotypic age. strengthened by recent study of 33 women, in which both
Interestingly, oxidative damage to mitochondrial O2 consumption and excretion of oxo8dG were measured.
membranes and proteins has itself been implicated in en- There was a highly significant positive correlation (P
hanced oxidant generation; exposure of isolated mito- 0.00007, r 0.64) between O2 consumption and excretion
chondria to the free radical generator 2,2-azobis(2-amino- of adducts (188). When the urinary output of the oxidative
propane)dihydrochloride or the cross-linking agent glutar- DNA repair products oxo8gua thymine glycol and thymi-
aldehyde resulted in mitochondria that were more able to dine glycol were compared in mice, rats, and humans,
generate H2O2 when fed exogenous substrate (308). When they were found to correlate with species-specific meta-
these in vitro studies are combined with the above evi- bolic rate (1, 287).
dence of increased oxidative damage, the picture that
emerges is a potential vicious cycle of oxidative damage
and oxidant generation. B. Antioxidant Defenses and Maximum
Life Span Potential

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


V. INTERSPECIES COMPARISONS
One version of the free radical theory proposes that
the differences in life span between species are due to
All of the data discussed in the previous section iden-
species-specific antioxidant capacity. Early work by Cut-
tify age-related differences within a given species. A com-
ler (57) tested this association: MLSP correlated positively
plementary, and in some ways more powerful approach,
with SOD but negatively with catalase and GPX (57). Re-
is to compare species that have different MLSP. Compara-
cently, other groups have revisited this hypothesis, with
tive biochemistry and physiology, inspired by the rate of
similar results. Comparisons between SOD, catalase, GPX,
living hypothesis, has played a key role in establishing
and GSH in brain, liver, and heart from mice, rats, guinea
the free radical theory. Several representative studies are
pigs, rabbits, pigs, and cows were carried out and revealed
described below.
that 1) SOD and catalase activities correlated positively
with MLSP, 2) GSH activity correlated negatively with
A. Oxidative Damage and Maximum MLSP, and 3) GPX correlated positively with MLSP in the
Life Span Potential brain and negatively in the liver and heart (309).
Interspecies comparisons between more distantly re-
The accumulation of cardiac lipofuscin in the monkey lated vertebrate classes found either no correlation or a
correlates with its cumulative O2 consumption, starting at negative correlation between antioxidants and MLSP. For
sexual maturation (226). Comparisons between rodent, instance, in the liver of fish, frogs, birds, and mammals,
dog, and primate species also revealed a close correlation strongly significant negative correlations were found for
between specific metabolic rate and lipofuscin accumula- catalase, GPX, and GSH, whereas all other measured anti-
tion (224, 225). The accumulation of the glycooxidation oxidants (SOD, GSH reductase, ascorbate, urate, GSH)
product pentosidine is similar. Short-lived species such failed to correlate with MLSP (190). When the same anti-
as the shrew (MLSP 3.5 yr) or rat (MLSP 3 yr), with oxidants were measured in brain and lung tissues of the
high specific metabolic rates, accumulate pentosidine in same species, virtually identical results were obtained
collagen at a rate of 0.30.5 pmol/mg annually. Longer- (12, 247).
lived primates such as monkeys and humans accumulate Comparisons between two very closely related spe-
pentosidine at about one-tenth as fast (284). cies of mice, the house mouse (M. musculus) and the
Interspecies comparisons of protein oxidation mirror white-footed mouse (P. leucopus), which lives twice as
these results. The protein carbonyl content of 15-day-old long, have revealed higher levels of SOD, catalase, and
individuals of five species of fly correlated negatively with GPX in brain and heart extracts of the latter, with the
mean life span; the longest-lived species (Drosophila mel- differences in GPX being the most dramatic. Also, when
anogaster, mean life span 65.5 days) had roughly one- brain homogenates of the two species were challenged
third the level of protein oxidation as shortest-lived spe- with X-irradiation, protein carbonyls accumulated at a
cies (Phaenecia sericata, mean life span 29.5 days) threefold higher rate in shorter-lived M. musculus than
(310). The content of protein carbonyls was 20% higher longer-lived P. leucopus (304).
in the heart, and 50% higher in the brain, of 3.5-mo-old Together, these results suggest that the evolution of
individuals of Mus musculus (MLSP 3.5 yr) than in 3.5- longevity has not been associated with a clear-cut in-
mo-old individuals of Peromyscus leucopus (MLSP 8 crease in antioxidant capacity, at least across the broad
yr) (304). sweep of evolution. (The comparisons between M. mus-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


562 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

culus and P. leucopus invite the speculation that the more five mammalian species and thoracic muscle mitochon-
recent radiation of long-lived species may be associated dria of two species of fly (311), and also found that longer
with the coordinate upregulation of antioxidant defenses.) MLSPs were associated with lower mitochondrial oxidant
In any case, what emerged from the comparative biochem- generation. The mitochondria from flies produced from
istry of free radical defenses is a fascinating paradox: 6-fold more to 300-fold more oxidants than those from
birds, which typically have much longer long life spans mammals. In a more recent comparison of heart tissues
than rodents, have lower activities of antioxidants and of eight diverse mammalian species of widely varying life
higher metabolic rates. This paradox has inspired the in- span, submitochondrial particles of the short-lived species
terspecies comparisons of mitochondrial oxidant genera- were found to generate more O20j than those of long-lived
tion discussed in section V. species, a property correlated directly with the concentra-
tion of CoQ9 (coenzyme Q possessing 9 isoprene units in
its isoprenoid tail) and inversely with CoQ10 (although
C. Generation of Reactive Oxygen Species and experiments intended to demonstrate a direct relationship
Maximum Life Span Potential between CoQ9:CoQ10 ratio and O20j generation failed to do
so) (178a).

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


The rate of living hypothesis identified an inverse Interestingly, it appears as if the concept of intrinsic
correlation between metabolic rate and life span, and the mitochondrial radical generation may also partially ex-
free radical theory supplied a convenient mechanistic plain longevity differences between more closely related
link: the mitochondrial generation of oxidants. Appealing species. Measurements of O20j and H2O2 generation by
as this theory is, one of the problems with the rate of isolated mitochondria from the mouse species M. muscu-
living hypothesis is the conspicuous lack of fit of a few lus and P. leucopus revealed that the former, which live
groups, notably birds and primates. Both of these groups half as long as the latter, also generate 4874% more
live longer than their specific metabolic rate would pre- O20j and 300500% more H2O2 . [As is the case with rats
dict. As a result, the total lifetime oxygen consumption of and pigeons, long-lived P. leucopus has the higher specific
these groups is quite a bit higher than other groups (e.g., metabolic rate of the two species (304).] Finally, a com-
during their MLSP, pigeons and canaries consume 465 and parison of five species of dipteran fly whose mean life
1,222 liters O2/g, respectively, whereas mice, rats, guinea spans vary by twofold also showed that the rate of genera-
pigs, and trout consume 77, 28, 48, and 26 liters O2 , respec- tion of mitochondrial O20j and H2O2 correlated negatively
tively) (247, 248). However, it cannot be assumed that the with life span (310).
generation of oxidants is a direct function of metabolic Together, interspecies comparisons of oxidative dam-
rate, since mitochondria of different species (or even tis- age, antioxidant defenses, and oxidant generation provide
sues) may exhibit different rates of intrinsic oxidant gen- some of the most compelling evidence that oxidants are
eration. These lines of reasoning by two laboratories have one of the most significant determinants of life span.
led to a similar line of experimentation, the results of
which suggest that longevity is associated with a lesser
capacity for mitochondrial oxidant generation (13, 172, VI. DIETARY RESTRICTION
173, 304, 310).
For example, isolated pigeon mitochondria from Limiting the dietary intake of mammals is a well-
brain, liver, or heart consume from two- to threefold as established way to extend life span (340). Interestingly,
much oxygen as isolated rat mitochondria from the same early expectations that dietary restriction (DR) would
tissues. However, pigeon mitochondria generate only one- lower metabolic rate per se have not been confirmed, and
third to one-half as much H2O2 as rat mitochondria under so if DR attenuates oxidative damage, it is not via a simple
the same conditions. As a result, the calculated percent- reduction in oxygen consumption (312, 339a). Rather,
age of O2 converted to O20j/H2O2 by mitochondria from there is now convincing evidence that dietary restriction
lung, liver, and brain is 10-fold lower in pigeons than in may act in part by decreasing oxidative stress and increas-
rats. This lower rate of oxidant generation corresponds ing antioxidant defenses and repair (352). For example,
with the roughly 10-fold longer life span of pigeons (13). mice whose caloric intake was restricted by 40% (DR)
A second, independent comparison of mitochondrial oxi- compared with ad libitum (AL) fed controls, and who
dant generation by pigeons and rats found similar results lived 43% longer on average, had a less rapid accumulation
(172). Detailed respiratory comparisons have shown that of protein carbonyls in brain, heart, and kidney. Moreover,
generation of oxidants from both complex I and complex the generation of H2O2 and O20j by mitochondria and sub-
III of the mitochondrial electron transport chain is lower mitochondrial particles from DR animals was lower than
in pigeon than rat (130a). from AL animals, and also increased less with advancing
In fact, the earliest interspecies study of mitochon- age. In this study, however, DR did not appear to enhance
drial oxidant production compared liver mitochondria of antioxidant defenses (305). In another study, 40% DR of

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 563

rats increased the activities of SOD, catalase, and GPX at markedly reduces their mean and MLSPs, and also in-
older ages; free radical damage, as measured by TBARS creases the rate of accumulation of protein carbonyls in
and lipofuscin accumulation, was correspondingly lower whole body extracts (299) and in isolated mitochondria
(253). In a third study, focusing on mitochondrial oxidant (303). Similarly, elevated atmospheric O2 decreased, and
generation and membrane properties of synaptosomal subnormal oxygen increased, the mean and maximum life
preparations, 40% DR reduced mitochondrial oxidant gen- spans of nematodes (138). Of course, the principal draw-
eration in both young and old samples and prevented the backs of this type of experiment are its inapplicability
age-dependent decline in membrane fluidity, despite the to mammals and the fact that elevations or decreases in
fact that increases in the cholesterol-to-phospholipid ratio ambient O2 in such species will likely be confounded by
were common to both groups of animals (43). Other ex- the overt pathology of hypo- or hyperoxia.
periments have been recently reviewed (335, 352).
Dietary restriction also delays cancer incidence,
which may be a reflection of fewer mutations induced by IX. SUPPLEMENTATION WITH DIETARY
oxidants in DR animals. Dietary restriction has been ANTIOXIDANTS
shown to strengthen DNA repair (111), although few stud-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


ies have focused specifically on repair of oxidative lesions. As one of the most intuitive approaches to testing
Not only caloric restriction, but also restriction of protein the free radical theory, nutritional supplementation has
intake, can forestall the occurrence of cancer in rodent been attempted with numerous species and compounds,
models. Although the mechanism by which protein re- with the result that mean life span has been extended
striction operates is not clear, enhanced protection in some instances (122). In mammals, results have been
against cellular degeneration, including oxidative damage, mixed. To cite recent studies, life-long oral supplementa-
is a likely candidate. It is therefore interesting that the tion of rats with UQ was without effect on mortality (189),
level of protein carbonyls in animals fed a low-protein whereas supplementation of the senescence accelerated
diet was significantly lower than in animals fed standard mouse strain (see sect. XII) with bCATECHIN, a com-
lab food and that treatment of the animals with g-irradia- mercial supplement with antioxidant activities (174, 349)
tion induced a greater increase in protein carbonyls in resulted in life extension (175). Negative results from the
high- than in low-protein animals (350). feeding of antioxidants have been rationalized by arguing
that many of the antioxidants fed to mammals interfere
with mitochondrial respiration, and so the failure of anti-
VII. MANIPULATION OF RATE OF LIVING
oxidant trials to extend MLSP may have been due to their
toxicity (127). In a recent short-term feeding experiment,
The metabolic activity of caged houseflies can be re-
in contrast, the thiol donor N-acetyl-cysteine was found
duced by limiting the space available for flight (low activ-
to markedly improve mitochondrial ETC complex activity
ity LA vs. high activity HA). It is expected that the
in rats (214) and short-term feeding with a complex anti-
LA flies should live longer due to lower O2 consumption
oxidant-containing plant extract attenuated oxidative
and generation of oxidants, and this is in fact observed.
damage in rat mitochondria (272a). Experiments aimed
LA flies exhibit more than a twofold increase in both mean
at extending life span should probably be preceded by
life span and MLSP. Measurements of protein carbonyls
such short-term trials, to document the uptake, distribu-
in 14-day-old LA and HA flies, which have not yet begun
tion, and biochemical action of the compounds.
to die, revealed an elevation of more than 50% in HA flies
Although no large-scale human nutritional interven-
relative to LA flies in both whole body extracts (299) and
tion trials have been aimed specifically at the study of
in mitochondria (303).
aging, a few have been undertaken in the hopes that di-
Under unfavorable conditions, the nematode C. eleg-
etary antioxidants might help prevent cancer. Three simi-
ans may undergo a morphological transformation to form
lar and heavily scrutinized studies are the a-tocopherol
a dauer larva, a resting stage that does not feed, exhibits
b-carotene (ATBC) prevention study (4), the physicians
altered metabolic activity, and can survive for several
health study (129), and the b-carotene and retinol efficacy
times the MLSP of adult nematodes before returning to
trial (CARET) (236), all of which were designed to assess
the adult stage. Interestingly, dauer larvae have also been
the effect of antioxidants on lung cancer risks. Although
found to have increased SOD activity (178).
the details of these enormous experiments are not im-
portant to this discussion, the overall result was instruc-
VIII. MANIPULATION OF OXYGEN TENSION tive: there was either no decrease or a modest increase
in lung incidence and mortality with b-carotene adminis-
Another experimental manipulation that modulates tration.
life span is growth under different O2 tensions. The growth These negative results, as well as negative results
of houseflies in an atmosphere of 100% O2 , for example, described above using laboratory rodents, should not be

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


564 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

taken as evidence that the free radical theory of aging is ise, not only for the study of aging, but also for its thera-
flawed. In fact, they prove merely that a complex organism peutic treatment, subsequent experiments reported that
like a human or rodent is unlikely to respond predictably the results were irreproducible (31, 32). Further follow-
to crude manipulations such as supplementation with one up work from a third laboratory confirmed the initial find-
or a small number of compounds (22), as well as that a ings in gerbil brain, but at the same time found no effect
single end point (such as lung cancer) is not equivalent of PBN in lowering the level of protein carbonyls in gerbil
to aging. Since the time when the results of these human heart or mouse brain (75). The chronic treatment of aged
trials were announced, a number of explanations have rats with PBN was found to reverse age-related cognitive
been forwarded to explain the paradoxical promotion of impairment (294). Finally, the administration of PBN from
cancer by b-carotene (252), which indicates that the ulti- age 20 wk throughout the life of a short-lived strain of
mate value of these trials may be a more precise under- mouse [the senescence accelerated mouse (SAM)] in-
standing of this antioxidant. Ultimately, what the entire creased its life span from 42 to 56 wk (an increase of
experience should teach the field of molecular gerontol- 33%) (73). As is argued (in sect. IX) to be the case for
ogy is that at least until the biochemistry of dietary and experiments with dietary intervention, maximum worth
cellular antioxidants is better understood, dietary trials in will be derived from pharmacological experiments when

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


laboratory animals or humans will remain unreliable tests they include not only end-point mortality measurements
of the free radical theory; molecular gerontology should but more focused studies of cellular effects.
focus on more instructive experiments.

XI. IN VITRO SENESCENCE AND OXIDANTS


X. ADMINISTRATION OF PHARMACOLOGICAL
ANTIOXIDANTS The growth of normal diploid fibroblast in vitro is
limited; after a specific number of population doublings
Dramatic antiaging effects have been observed by (PDs) characteristic of the species, tissue, and age of the
Floyd and co-workers (234) with the free radical spin trap donor (the Hayflick limit, named for the phenomenons
N-tert-butyl-a-phenylnitrone (PBN). Initial studies used discoverer), the cells cease replication, which culminates
PBN as a spin trap to measure radical generation during in a viable, nondividing population with characteristic se-
ischemia-reperfusion injury of gerbil brain (234). During nescent morphology (56). Research on such replicative
reperfusion, the appearance of protein carbonyls corre- senescence, and the vigorous debate about its relevance
lated well with the loss of glutamine synthetase activity to organismal aging, have been recently reviewed (30, 55,
and the appearance of PBN-dependent nitroxide radical. 56). For the purposes of this discussion, a few points
Interestingly, it was also discovered that pretreatment are worth reiterating. 1) There are numerous correlations
with PBN attenuated protein oxidation and loss of gluta- between organismal MLSP and replicative senescence. Fi-
mine synthetase activity and reduced lethality. broblasts from species with shorter MLSP exhibit a
In a subsequent study that investigated the effect of shorter in vitro life span than cells from more long-lived
PBN on spontaneous protein oxidation, it was discovered species. Similarly, fibroblasts from older individuals pos-
that twice-daily administration of PBN to old gerbils (32 sess a shorter in vitro replicative potential than those from
mg/kg) reversed age-related oxidative damage (34). For younger individuals (30), although this generalization has
example, during normal aging of gerbils, the level of brain been questioned (55). Moreover, cells from patients with
protein carbonyls increased 185%, and glutamine synthe- Werners syndrome, a disease of accelerated aging (see
tase activity declined by 35%. A marked drop in neutral sect. XVII), also complete fewer PDs before crisis, when
protease activity was also observed. Old animals adminis- replication ceases (30). 2) In vitro senescence may be
tered PBN for 2 wk, however, experienced a reversal of relevant to organismal life span in at least two ways. First,
these trends: protein carbonyls dropped, and glutamine as one of the likely components of tumor suppression,
synthetase and neutral protease activities were virtually cellular senescence is necessary (but not sufficient) for
restored. This effect was reversible, since over a 2-wk extended life span. In other words, a long life span (as in
period after cessation of PBN treatment, the oxidation of humans, for instance) requires the absence of cancer at
proteins increased, and the activities of glutamine synthe- younger ages, and this is achieved in part by growth arrest
tase and neutral protease fell. Most impressive, however, of preneoplastic cells by mechanisms that operate during
was the performance of the PBN-treated old gerbils in a in vitro senescence (30, 269). Support for this assertion
test of short-term memory; whereas the old gerbils were are the observations that fibroblasts from rodents (which
twice as likely to make errors as young gerbils, the old have a significant cancer incidence after 23 yr of age)
animals treated with PBN performed as well as the spontaneously immortalize in vitro, whereas fibroblasts
young (34). from humans (who are relatively free of cancer until much
Although these preliminary results held great prom- later ages) do not (30), and that immortalization is associ-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 565

ated with inactivation of p53 and Rb, which are frequently and stimulates lipofuscin accumulation (333). Direct feed-
mutated in neoplasms (107). Second, senescent cells ap- ing of synthetic lipofuscin particles to fibroblasts, which
pear to occur in vivo, as judged by enzymatic markers phagocytose and accumulate the material, simulates se-
characterized in vitro (68). It has been suggested that even nescence (333). Most intriguing, perhaps, is the revelation
at a low frequency, such cells may exert a disproportion- that 40% O2 culture resulted in a rapid loss of telomeric
ate deleterious effect on tissue physiology by dominant DNA as well as other cytogenetic alterations characteris-
effects of aberrant cellular regulation (e.g., expression of tic of senescent cells, and a permanent loss of replicative
inappropriate cytokines) (30). 3) In vitro senescence has capacity reminiscent of the growth arrest of cells with a
been closely associated with the shortening of telomeric pulse of H2O2 (330a, 334). The measurement of DNA sin-
DNA, at least in humans (56). In human somatic cells, the gle-strand breaks in telomeric sequences of such cells
average length of telomeric DNA shortens by 50 bp/PD revealed an increase under hyperoxia (334), confirming
in vitro, and by 15 bp/yr in vivo (30). the idea (40) that DNA damage by oxidants may be in-
Leaving aside the question of the relevance of replica- volved in replicative senescence in vitro and extending
tive senescence, is there any evidence that oxidative dam- the model to embrace the theory of telomere shortening
age contributes to it? As early as the mid 1970s, this hy- (334). Interestingly, it has also been recently shown that

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


pothesis was tested, by culturing human fibroblasts under the repair of ultraviolet light-induced pyrimidine dimers
conditions of low O2 partial pressures and supplementing in telomeric DNA is lower in fibroblasts from an old donor
culture medium with the antioxidant a-tocopherol, with than a young donor (171).
somewhat equivocal results (10, 241, 242). During the Together, these experiments are consistent with the
1980s, a handful of studies documented the increased sen- hypothesis that oxidative DNA damage, acting to shorten
sitivity of late-passage cells to oxidant challenge (139, 140) telomeric sequences and thereby induce DNA damage-
and noted that hyperoxic exposure could shorten in vitro responsive growth arrest controls (107), plays a role in
life span (140). Later studies, looking to PD-dependent in vitro senescence. Two recent studies permit a further
changes in antioxidant defenses as a potential explanation speculation. It has been shown that overexpression of the
for the increased sensitivity of late-passage cells to oxi- oncogenic ras gene in primary cells, which initially serves
dants, failed to account for these observations (250), al- as a strong mitogenic stimulus, rapidly induces premature
though an impaired ability of late-passage cells to reini- senescence in a p53- and p16-dependent fashion (286).
tiate DNA synthesis after hyperbaric O2 exposure was Interestingly, it has also been recently reported that ras
established (141). More recently, a loss of oxo8gua glyco- induces cellular oxidant generation (147). It is tempting
sylase activity with increasing passage number has been to suggest that the aggressive mitogenic stimulus (286)
reported (133), suggesting that decreased repair may be of oncogenic ras, credited with inducing senescence, may
involved. be oxidative in nature. There may be a role for oxidants
Recently, the issue has been revisited, and it was in replicative senescence.
shown that a short (2 h) exposure to H2O2 (200 mM) in-
duces an abrupt senescent-like growth arrest in human
XII. CLASSICAL AND POPULATION GENETICS
diploid fibroblasts (40). Intriguingly, cells that resumed
cell division were found to have a markedly attenuated
In comparison to the measurement of age-related
life span; a short-term oxidative stress had resulted in a
changes, genetic approaches have the potential to provide
long-term loss of PDs. The inhibition of H2O2 growth ar-
more definitive information about aging, by supplying
rest by a metal chelator suggested that a Fenton-derived
model organisms (mutant strains with altered life span)
jOH radical was involved, and it was proposed that oxida-
that differ at a small number of loci. Many of the geneti-
tive DNA damage and its sensing by cell cycle checkpoint
cists favorite organisms have been found to be amenable
controls may be involved in the phenotype (40). Indeed,
to studies of aging and have furnished evidence of the
measurements of DNA oxidation indicated that senescent
role of oxidants (196, 289).
cells suffer greater oxidative DNA damage than young
cells and that the addition of the antioxidant PBN to the
culture medium prolonged the life span in a dose-depen- A. Caenorhabditis elegans Genetics
dent fashion (41). In these and other studies, it was shown
that culture under low oxygen tension extended fibroblast The emergence of C. elegans as a convenient meta-
life span (41, 355) and that the inhibition of catalase de- zoan genetic model has attracted scientists seeking to
creased it (355), implicating endogenous H2O2 in replica- discover genes controlling life span, or genontogenes
tive senescence. (152, 158). Several long-lived mutant strains of C. elegans
Interestingly, lipofuscin accumulates in fibroblasts in have been identified, including age mutants (the originally
vitro as well as in vivo, and culture under mildly hyperoxic identified phenotype being extended life span), daf mu-
conditions (40% O2 partial pressure) decreases life span tants (identified by abnormalities in the formation of

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


566 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

dauer larvae), and clk (clock) mutants [which exhibit a biosynthesis of the mitochondrial electron transport com-
pleiotropic behavioral deceleration (slower feeding, defe- ponent UQ (252a). It has been proposed that the deceler-
cation, and movement)] (158). These two classes of mu- ated metabolic rate in Clk-1 mutants of C. elegans may
tant have collapsed into a single category with the cloning result in a correspondingly slower rate of oxidant genera-
of age-1 and its identification as an allele of daf-23 (219). tion (127b), although the authors of this work note that so
Strains of age-1 live 65% longer on average and have a far data are merely consistent with such a proposal.
MLSP 110% longer than wild type (153). The trait is reces- A mutant that is in some ways the opposite of age-1
sive, and biochemical studies have revealed that strains resulted from a screen for sensitivity to O20j in a chemi-
carrying age-1 alleles have enhanced oxidative defenses. cally mutagenized population of C. elegans. The resulting
For example, when the sensitivities of wild-type and age- mutant, mev-1, is hypersensitive to both paraquat and
1 strains to H2O2 were compared, it was found that the hyperoxia. The activity of SOD in mev-1 was found to be
50% effective lethal dose (LD50) of the wild-type remained roughly one-half of wild type, and the average life span
constant, whereas the LD50 of the age-1 mutant increased was reduced by 35% (148). Lipofuscin-like fluorescence
with age, an indication of increasing resistance to oxi- accumulated more rapidly in mev-1 mutants that in wild-
dants (178). Similar comparisons showed that age-1 mu- type animals (144). The life span of both mev-1 and wild-
tants are more resistant to O20j as well (330). In both

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


type nematodes was lengthened under hypoxia and short-
cases, greater resistance to oxidants is associated with ened by hyperoxia, but the effect was more dramatic for
elevated activities of SOD and catalase. Whereas in wild- the mutants (138).
type animals these activities remained constant through-
out the life span, in age-1 strains they increased with B. Drosophila Genetics
advancing age, such that the difference between the mu-
tants and the wild type grew increasingly pronounced at According to the evolutionary concept of antagonis-
old ages (178, 330). In addition to being resistant to oxi- tic pleiotropy, natural selection favors traits that benefit
dants, age-1 strains have increased heat tolerance, so it is organisms during their period of reproduction. Therefore,
likely that enhanced ability to cope with all environmental by imposing a strong artificial selection that consistently
stresses, including oxidative ones, is the result of the loss- rewards females capable of late reproduction, one ought
of-function mutation (185). to be able to select against those pleiotropic traits that are
A comparative study of daf-2 and age-1 mutants re- most antagonistic to long-term survival. After a suitable
vealed that they lie in a common pathway. Life extension number of populations, a long-lived phenotype should
by either requires the action of wild-type daf-16, and their emerge from a large, heterogeneous population.
effects on life extension are not additive (72), and so it In the early 1980s, Rose (264) followed precisely this
was reasoned that they work in tandem. Their cloning re- strategy using Drosophila thereby validating the concept
vealed that daf-2 is a member of the insulin receptor family of antagonistic pleiotropy and also providing five indepen-
(160a) and that age-1 is a catalytic subunit of phosphatidyl- dently selected lines (referred to as lines O1-O5) for fur-
inositol 3-kinase (219), suggesting that together they trans- ther experimentation. The strength of this model is that,
duce an insulin-like metabolic signal with pleiotropic ef- unlike age-1, lines O1-O5 are not single gene mutants, but
fects on metabolism, stress resistance, and dauer formation rather populations in which any number of allelic changes
(127b, 291a). The more recent cloning of daf-16, revealing will have occurred. Hence, biochemical alterations that
that it is a member of the hepatocyte nuclear factor 3 (HNF- are found to have arisen independently in many of these
3)/forkhead family of transcriptional regulators (183a), has lines are likely to reflect the type of complex polymor-
provided insight into the entire pathway, since in humans, phisms that confer increased life span to long-lived spe-
insulin signaling antagonizes HNF-3. The implications, cies such as Homo sapiens.
then, are 1) that signaling through daf-2 and age-1 ulti- Therefore, it is significant that in a comparison of the
mately may serve to antagonize the control of a genetic lines O1-O5 with lines B1-B5 (control lines not subjected
program controlled by daf-16, and 2) that this pathway to selection), all of the long-lived populations had ac-
may be relevant to human aging, since it appears to have quired a higher frequency of a relatively rare SOD allele
been conserved (183a). Of course, the connection of these that is associated with higher in vitro activity (328). In
proteins to the aerobic respiration and the rate of living, via Drosophila, there exist two electrophoretic alleles of
insulin-mediated metabolic control, has escaped no ones SOD, designated F (with lower in vitro activity) and S
notice (183a). (with higher in vitro activity). Whereas the genotypes of
The cloning of the clk-1 gene also implicates respira- all B flies tested were FF (no S alleles were detected),
tion rate in C. elegans life span, albeit indirectly. The gene both FS and SS individuals existed in all five O lines, such
is homologous to a yeast gene called CAT5/COO7 (79a), that the average allele frequency of S ranged from 0.10 to
which is necessary for the genetic control of the shift from 0.28 [0.25 { 0.05 (SE)]. In all five independent instances
anaerobic to aerobic metabolism (127b), as well as the of laboratory evolution, SOD activity had increased.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 567

C. Rodent Genetics including humans (17, 192, 266), which forcefully makes
the case that oxidants are significant mutagens in vivo.
The senescence-accelerated mouse (SAM) is an in- There is, moreover, plentiful direct evidence of oxidative
creasingly important model in gerontology (244); the se- mutagenesis in eukaryotic cells, such as the demonstra-
nescence-prone strain (SAM-P) has a mean life span of 9 tion that disruption of the metallothionein gene in tissue
mo, compared with 13 mo for a senescence-resistant culture, which may abrogate the cells ability to chelate
(SAM-R) strain. In liver tissues, the activity of mitochon- redox active metals such as iron, results in a 5- to 10-fold
drial SOD in the SAM-P strain was about one-half that in increase in spontaneous mutagenesis (267). Also, muta-
the SAM-R strain, a result that was consistent at all ages genesis in cells from humans with Fanconis anemia, a
examined (244), which could explain the more rapid age- condition associated with oxidative stress, displayed a
related increase in lipid peroxidation seen in this tissue higher mutation frequency than controls and a marked
(244) as well as independent reports of enhanced oxida- increase in mutation frequency in response to the eleva-
tive stress in SAM-P mice (186). tion of O2 partial pressure (183).
A potentially very useful animal model for testing the Despite these results, the proportion of mutations
free radical theory of aging is the S strain of Wistar rat, due to oxidants in the presence of wild-type antimutators

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


selected for sensitivity to cataractogenesis by galactose. is still an unknown quantity. Recent experiments have
A heritable increase in cellular hexose uptake by cells illustrated that dietary supplementation with antioxidants,
from the S strain is associated with an increased intracel- which reduces irradiation-induced mutagenesis, appears
lular generation of oxidants, increased endogenous lipid to reduce the spontaneous mutation rate as well (99, 100),
peroxidation, mitochondrial dysfunction, an increased suggesting that antioxidant scavenging in vivo is incom-
age-specific incidence of degenerative diseases, and ear- plete. The quantitative contribution of oxidants in somatic
lier aging than the galactose-resistant R strain (271, 272). mutagenesis is an open (and important) question.
Moreover, SOD and catalase activities in the blood of S
rats are one-half that in R rats, which may have explained
the increased in protein oxidative damage that was also B. Oxidants and Mitochondrial Somatic Mutations
seen in the S strain (345).
A burst of activity was stimulated by the report that
a large deletion of the mitochondrial genome (DmtDNA)
XIII. MOLECULAR GENETICS
accumulates exponentially with age in humans and by the
speculation that this may be the work of mitochondrial
A. Oxidants and Nuclear Somatic Mutations oxidants (53). Since this first detection of DmtDNA in
normal elderly humans, dozens of studies have confirmed
Oxidants are twice removed from the somatic muta- the finding, detecting the same common deletion in hu-
tion theory of aging, since two requirements need to be mans as well as different deletions in other species, in-
satisfied for the relevance of oxidants to be ensured: 1) cluding C. elegans (206), mice (44, 323), rats (76), and
oxidants must account for a significant proportion of mu- rhesus monkeys (179). Increased frequencies of DmtDNA
tations, and 2) mutations must account for a significant in brain samples have been associated with Huntingtons
proportion of the phenomena of aging. The second crite- disease and AD, degenerative diseases of aging that affect
rion is complex and has been recently discussed (71a, millions of people (48, 143).
195, 218, 338b); suffice it to say that to the extent that In most of these cases, a significant accumulation of
cancer is a degenerative disease of age (6), there is little the deletions occurs only at advanced ages and is associ-
doubt that mutations are relevant. What remains to be ated primarily with postmitotic tissues. This mosaicism
seen is whether or not the age-related degenerative of mtDNA deletion (313), which is reminiscent of the pat-
changes that do not involve neoplasia are also the result tern the maternally transmitted mitochondrial diseases, is
of mutations (79) (or epigenetic alterations; Ref. 136). hypothesized to be partly due to the lack of proliferative
With the assumption that they are, the question then competition in postmitotic tissues (50, 336). Whereas in
remains, To what extent are spontaneous mutations due dividing tissues cells that inherit a relatively greater per-
to oxidants? Some of the best evidence so far of the rele- centage of DmtDNA molecules by unequal mitotic assort-
vance of oxidative mutagenesis in vivo is an argument by ment stand to compete less well in subsequent rounds of
analogy. In E. coli, mutation rates can be elevated from cell division, this phenotypic sieving will not operate in
10- to 1,000-fold by the loss of mutator genes mutM, mutT, a nondividing tissue.
and mutY, whose gene products are involved in repair There are reasons to believe that the deletion of
and prevention of oxidation of guanine in DNA (210). mtDNA may stem from oxidative damage (51). For one, the
Significantly, homologous genes and novel genes with an- steady-state burden of oxidative adducts in mtDNA has
alagous activities have been found in a number of species been measured to be 16-fold higher in rat mtDNA than in

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


568 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

nuclear DNA (16, 260). Moreover, quantitative polymerase genetically engineer long-lived Drosophila by bolstering
chain reaction analyses of DmtDNA accumulation in differ- antioxidant defenses have been successful. The first such
ent areas of the brain have consistently revealed highest efforts involved the introduction of a bovine cDNA for
frequencies of areas of high metabolic activity (47, 54). De- Cu,Zn-SOD into Drosophila, which resulted in a signifi-
spite the promise of these results, it is not self-evident that cant increase in SOD activity as well as resistance to oxi-
the observed accumulation of DmtDNA represents more dative stresses and a small, statistically significant in-
than a biomarker (182). For one, the highest frequencies of crease in mean life span (85). Whereas a repeat of this
DmtDNA detected, even at the oldest ages and in the most procedure with the Drosophila Cu,Zn-SOD gene resulted
deletion-prone tissues, are generally no higher than 0.1%. in only marginal effects on resistance to oxidants, and
Because in maternally transmitted mitochondrial genetic failed to increase life span, the introduction of single cop-
disorders symptoms are often absent even when more than ies of Drosophila SOD and catalase resulted in a number
one-half of the mtDNA molecules carry deletions (336), it of strains with significantly extended mean and maximum
is hard to see how such low frequencies could exert an life spans (237).
effect. On the other hand, because there may be numerous As expected, the augmentation of antioxidant activi-
deletions existing independently, measurements of a single ties in the transgenic flies led to significant improvements

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


one may merely represent the tip of the iceberg of mtDNA over controls in numerous age-related indexes of oxida-
damage (54). The use of long polymerase chain reaction tive metabolism: 1) the accumulation of protein carbonyls
to amplify the entire mitochondrial genome has revealed the was reduced; 2) the rate of oxygen consumption at middle
presence of numerous mtDNA rearrangements in human age and old age was higher; 3) the 50% loss of G-6-PD
skeletal muscle (207). On the other hand, a recent attempt activity with age was reduced to 10%; 4) the age-related
to measure an increase in somatic point mutations in human doubling of the load of oxo8dG was virtually abolished,
skeletal muscle was unsuccessful (243). as was the roughly twofold increase in sensitivity to the
The analysis of progressively smaller bundles of cells induction of oxidative DNA damage; and 5) the age-re-
has shown that the DmtDNA frequency is unevenly dis- lated threefold increase in mitochondrial oxidant genera-
tributed (282). For example, when a muscle homogenate tion was reduced by about one-half (237, 298). In short,
of rhesus monkey fibers is analyzed, the detected fre- the predictions of the free radical theory were confirmed.
quency of DmtDNAs is 0.020.1%. However, restriction
of the sample size to 10-fiber bundles decreases the num-
ber of positive reactions but increases the measured fre- B. Transgenic Mice
quency of DmtDNAs in positive reactions to 4.613.2%.
In other words, a frequency of 0.1% could represent 100% Ultimately, transgenic mice are the molecular geron-
deleted molecules in 1 cell of 1,000 (180). At any given tologists Holy Grail, and it is not until similar experiments
point in time, only a small number of cells may carry are performed in mice that the free radical theory will
deletions, but if these cells continually die as a result of have been satisfactorily tested. Although such experi-
a failure of oxidative phosphorylation, and other deletions ments will clearly be exciting, work that has already been
arise de novo, then the low steady-state frequency of dele- performed with mice transgenic for SOD suggests that
tions could obscure a high rate of DmtDNA-associated their phenotypes may not be straightforward. As has been
cell death. stressed, antioxidant defenses are interactive and coordi-
Finally, it has been proposed that a vicious cycle of nately regulated, and it has become apparent that oxidants
increasing oxidative damage and stress of the type dis- play important roles as signaling molecules (286). As a
cussed in section IIF is particularly likely in the case of consequence, up- or downregulating a single antioxidant
mtDNA, since it encodes components of the mitochon- may disrupt an internal balance and either exacerbate
drial electron transport chain (52). This hypothesis is sup- oxidative stress or otherwise compromise cellular physi-
ported by the fact that in statistical terms, a random muta- ology. Indeed, the phenotype of Cu,Zn-SOD-overex-
tion of mtDNA is most likely to affect complex I, com- pressing mice appears to mimic the pathology of Downs
bined with the observation that a complex I poison syndrome (for which they may be a model, due to the
(MPTP) triggers neuronal cell death in mice (52). It re- localization of Cu,Zn-SOD to the trisomic chromosome
mains to be shown that known somatic mtDNA mutations 21; Ref. 36). Moreover, it has been recently discovered
do, in fact, result in increased oxidant generation. that a complete knockout of Cu,Zn-SOD has a phenotypic
effect only on a small subset of motor neurons thought
XIV. TRANSGENIC ORGANISMS to be involved in amyotrophic lateral sclerosis (255). Even
more surprising, the recent elimination of the enzyme
A. Transgenic Drosophila
GPX in a transgenic knockout mouse model resulted in
The modern acid test of a biological theory is the generally normal, fertile mice without overt signs either
generation of a transgenic organism, and recent efforts to of oxidative stress or increased sensitivity to oxidants

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 569

(134). Together, these results indicate that the promising recent work by two groups has independently revealed
results with Drosophila may be difficult to replicate, al- an artifact in the unusual protocol used for mtDNA extrac-
though no matter what the results, antioxidant overex- tion in these experiments: a nuclear pseudogene, rather
pressing strains will clearly be extremely useful. By cross- than mtDNA itself, was amplified, and so the conclusions
ing animals with different transgenes, it will be possible from these studies are invalid (131b, 336a).
to dissect antioxidant interactions and their effects on
aging.
XVI. INHERITED DEGENERATIVE DISEASES

XV. SPORADIC DEGENERATIVE DISEASES Werners syndrome (WS) is a segmental progeroid


syndrome of late-onset accelerated aging, associated with
an increased incidence of cancer, attenuated replicative
In the form of the specific age-related diseases in
senescence of cells in vitro, and, interestingly, an eleva-
elderly humans, there exists a natural laboratory of oxida-
tion of oxidized protein in isolated fibroblasts (315, 354).
tive stress. Indeed, investigation of the etiologies of hu-
The cloning (354) of the gene responsible for the syn-
man degenerative disease implicate oxidative stress in
drome (WRN) does not suggest an oxidative etiology,

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


cancer (5, 6), atherosclerosis (112, 343), diabetes (326),
however, as the protein product WRN appears to be a
and neurodegeneration (15, 18, 293), to name a few.
helicase (106a). WRN is homologous to the RecQ helicase
For example, the role of oxidants in AD has been a
of E. coli, which is involved in unwinding DNA during
particularly active area of research in recent years. It has
repair; hence, it has been proposed that the loss of heli-
been known for some time that the formation of amyloid
case activity in WS results in a deficiency in recombina-
deposits, a hallmark feature of the clinical diagnosis, is
tion-mediated repair in humans cells (106a, 291a). Ineffi-
associated with aberrant processing and folding of a pro-
cient or dysregulated DNA repair, in turn, may cause the
tein called amyloid b-precursor protein (AbPP) to form
chromosomal instability, which is one of the diseases
the amyloid plaques (Ab) characteristic of the disease
hallmarks and which may be the root of some or all of
(283). In recent years, it was discovered that not only
the diseases associated with the syndrome (106a, 291a).
the in vitro formation of synthetic Ab from AbPP can be
Although there is consequently no obvious causal link
accelerated by the presence of oxygen but also that Ab,
between mutations in WRN and oxidative damage, the
which is cytotoxic when fed to neurons in vitro, appears
increased protein oxidation in WS fibroblasts (315) allows
to stimulate the generation of oxidant (293). In fact, amy-
the speculation that measurable oxidative damage may
loid peptides themselves generate oxidants in vitro (130).
result from inefficient recombination-dependent repair,
As discussed above, it has also been recently discovered
chromosomal instability, and an ensuing disruption of ge-
that a cellular receptor (on microglial cells and endothe-
netic programs. If this turns out to be the case, the oxida-
lial cells) responsible for transducing the cytotoxic signal
tive phenomenology of WS will serve as an important
from Ab and inducing oxidant generation is none other
reminder that the measurement of oxidative damage does
than RAGE (344). Amyloid plaques have also been found
not imply oxidative causation! [On the other hand, if nor-
to bind redox-active iron, providing another potential pos-
mal aging, like WS, results in part from chromosomal in-
itive feedback loop for oxidative damage (292a). Although
stability, it may be that oxidative genotoxicity plays a
we do not have space to go into the details, what is rele-
central role (sect. XIIIA). In other words, it may be that
vant here is not so much the specifics but the realization
oxidative damage may both result from and result in chro-
that in numerous ways, oxidants act as effectors in the
mosomal instability.]
progression of the disease. In other words, research into
the etiology of AD (193b, 203a, 292b) [and also Parkinsons
disease (15, 44a, 145a), amyotrophic lateral sclerosis XVII. EPIDEMIOLOGY OF OXIDANTS
(346), and dozens of other age-related disorders] is reveal- AND ANTIOXIDANTS
ing that cytotoxicity very frequently involves oxidative
stress. Nutritional epidemiology has suggested that dietary
Even if oxidants turn out to be essential effectors of antioxidants are crucially involved in the prevention of
AD pathology, can they initiate the process? Recently, it degenerative disease (22), and global epidemiological
was claimed that a higher degree of heteroplasmy of studies have pointed to geographical and economic anti-
mtDNA is associated with individuals with AD versus con- oxidant deficiencies in their localized occurrence (6). For
trols (62), consistent with a genetic predisposition based instance, the loss of immune function with aging is amena-
on mitochondrial inheritance. What is intriguing in this ble to therapeutic treatment with antioxidants (208a); the
work is the reported association of these AD polymor- fact that both antioxidant status and immunological de-
phisms with higher intracellular oxidant generation by cline vary between individuals (209) suggests that there
mitochondria isolated from the patients (62). However, may exist heritable difference in antioxidant capacities,

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


570 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

linked to differences in immune function. To date, how- differences between long- and short-lived species have
ever, the epidemiological study of senescence is merely been documented in all components of oxidative stress.
a theoretical possibility and awaits powerful tools before Perhaps most instructive, however, have been results that
it can be affordably attempted. initially appeared contradictory, such as the fact that birds
are an outlier group in rate of living calculations. Attempts
to account for the discrepancy have led to promising ad
XVIII. SUMMARY
hoc hypotheses about the role of mitochondrial oxidant
generation in aging. Unfortunately, relatively few re-
A. Are Oxidants Responsible for Aging? searchers are familiar with or prepared to handle a variety
of experimental animals, despite the research potential of
We have outlined what we see to be the major experi- a diverse menagerie.
mental approaches to testing the free radical theory of In contrast, the model of dietary restriction has ap-
aging. In Table 1, the strengths and weaknesses of differ- peal precisely because it is so well established, homoge-
ent modes of experimentation are assessed. How power- neous, and relevant to specific human diseases (sect. VI).
ful are different types of experiment? Has the free radical Although early predictions that DR would result in a lower

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


theory been supported? metabolic rate have proved unfounded (312), the discov-
Before a discussion of the merits of individual experi- ery that oxidative defenses are enhanced by DR have
ments, it should be stressed that if one is to judge them strengthened both the free radical theory and the concept
by their abilities to falsify the free radical theory, they are of antagonistic pleiotropy (158). Although, currently, re-
all bound to fail, for the simple reason that the free radical sults from DR studies are merely consistent with the free
theory is very hard to falsify. The absence of a predicted radical theory, future experiments may reveal specific
outcome (for instance, life span extension by antioxidant pathways whereby DR upregulates defenses and repair.
supplementation) may often be explained by (justifiable) Manipulation of metabolic activity and oxygen con-
ad hoc reasoning. The physiology of oxidative stress, be- centration are interesting systems, if of somewhat ques-
ing a complex interaction of endogenous and exogenous tionable relevance to normal aging (sects. VII and VIII). One
factors, generally permits the explanation of negative drawback of such experiments is that they have generally
results. How then shall one judge the different types of been limited to invertebrates; the manipulation of metabo-
experiments and the theory itself? In fact, as the results lism in mammals is less straightforward. In the future,
discussed above illustrate, the momentum gathering be- transgenic mouse models may allow a specific manipula-
hind the free radical theory is not due to any single experi- tion of metabolic rate in a consistent genetic background.
ment or approach, but rather derives from the extraordi- Attempts to manipulate life span with nutritional and
narily multidisciplinary nature of current research. Al- pharmacological antioxidants obviously hold great ap-
though no single line of reasoning alone permits definitive peal, as direct and intuitive tests of the theory (sects. IX
conclusions, together they present a compelling case. and X). Although until recently they have not quite lived
[This philosophy, that the study of aging should employ up to their promise (at least in mammals), this may have
a combination of different approaches, has been convinc- been due to a lack of the appropriate compounds. The
ingly espoused before (83).] serendipitous discovery of PBNs pharmacological prop-
Age-related oxidative phenomenology continues to erties appears to have opened up new avenues.
provide evidence that senescence is associated with in- The study of fibroblast aging in vitro might appear
creased oxidant generation, a decline in the robustness the least likely area of research to support the free radical
of defenses and repair, and an accumulation of end prod- theory, being a phenomenon of replication per se, rather
ucts of oxidative damage (sect. IV). Although these trends than of chronological aging (sect. XI). Nevertheless, the
are only correlations, the study of age-related changes has results discussed above not only suggest that oxidants
been rewarded with surprises such as the receptor for may play a role, but they also have led to novel hypotheses
AGE (RAGE), which turns out to be involved in the etiol- about the dynamics of telomeric DNA in vivo. Future work
ogy of AD and evidence that lysosomal lipofuscin may in this area should be exciting.
directly disrupt cells in vivo and in vitro. In fact, oxidative The utility of genetics and transgenic animals in the
phenomenology has focused efforts on understanding the study of aging, on the other hand, does not need elabora-
interplay between the components of oxidative stress out- tion (sects. XII and XIV). At present, the chief limitation to
lined in Figure 1. The search for evidence of age-related genetic studies is that fact that the adult forms of C. eleg-
DNA and protein oxidation has stimulated research into ans and Drosophila are postmitotic. Molecular genetics
DNA repair, proteolytic salvage pathways, mitochondrial (sect. XIII), which in the study of aging has primarily cen-
defenses, and so on. tered around the concept of somatic mutagenesis, is a
Interspecies comparisons have better potential di- field with particularly powerful tools that are generally
rectly to test the free radical theory (sect. V). As predicted, applicable to any species. It is therefore very well suited to

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 571

comparative studies, as illustrated by the hunt for mtDNA the chemical previously known as endothelium-derived
deletions in a wide range of species. Moreover, because relaxation factor, is generated enzymatically by isozymes
mutational spectra provide information about the primary of nitric oxide synthase and is involved in vascular tone,
mutagenic event (256), molecular genetic studies may find innate immunity, and neuronal signal transduction. More-
the smoking gun of oxidative mutagenesis. over, NO reacts with O20j to form peroxynitrite (ONOO0),
Finally, the push to understand human sporadic and itself a powerful oxidant. Therefore, not only must nitric
inherited diseases through biochemical, genetic, and epi- oxide synthase be seen as a potential source of damaging
demiological methods may ultimately provide the most oxidants (249, 281), but O20j may now be considered a
detailed information about age-related oxidative stress, as physiologically relevant scavenger of the free radical sig-
is becoming apparent in the cases of cancer, atherosclero- naling molecule NOj (58).
sis, and neurodegenerative disorders (sects. XV XVII). The
complete sequencing of the human genome, and modern
C. Oxidants and Apoptosis
high-throughput genetic methods, are poised to provide
an unprecedented amount of genetic information about a
Another topic we have decided to omit, because of
biological sample of many millions of individuals whose
space limitations and because this is an extremely fast-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


life histories (medical records) are accessible. For in-
moving and frequently reviewed area, is the role of oxi-
stance, an approach to the genetic study of aging that is
dants and mitochondria in apoptosis. Suffice it to mention
only now attracting mainstream interest is the study of
that in the previous year there have appeared dozens of
human longevity outliers, or centenarians (292). Techni-
experiments documenting either the loss of mitochondrial
cal advances in the identification and analysis of polymor-
membrane potential (355a), the release of mitochondrial
phisms, fueled by the Human Genome project, may permit
cytochrome c (255a) [or an apoptosis inducing factor
the study of oxidative stress-associated alleles in these
(355a)] into the cytosol, or the generation of oxidants by
extraordinary individuals. An example of such an ap-
mitochondria during apoptosis (355a). Combined with the
proach is the identification of a coding mtDNA polymor-
mitochondrial localization of the Bcl-2 family of anti-
phism (a C-to-A transversion at mtDNA nucleotide posi-
apoptosis proteins, and copious evidence that oxidant
tion 5178 in the NADH dehydrogenase subunit 2 gene) in
challenges can induce apoptosis (74a, 154, 290a), these
a larger percentage of centenarians (62% with 5178A) than
experiments have shown that mitochondria are central to
controls (45% with 5178A) (322a). In this study, it was
the process of cell suicide (255a, 355a). Despite this gen-
also found that as the age of randomly selected hospital
eral consensus, however, a number of disagreements exist
inpatients and outpatients increased, the percentage car-
about the details of the mitochondrions role in the pro-
rying the 5178C polymorphism increased, suggesting that
cess (149a) and even greater uncertainty about the role
such individuals are at higher risk of degenerative disease.
of oxidants. An equally complex problem will be defining
The authors hypothesize that the 5178A polymorphism,
the role of apoptosis itself in aging (338a). For instance,
which is rare in all human populations except the Japa-
by eliminating neuronal cells, apoptosis may contribute to
nese, may be responsible both for the longevity of Japa-
neurodegeneration, yet effective apoptosis is also clearly
nese centenarians and, therefore, for the general longevity
critical in preventing (age-related) tumorigenesis. In
of the Japanese population as a whole. They suggest that
short, both the machinery and impact of apoptosis are still
such a polymorphism, affecting the mitochondrion, may
incompletely understood; conclusions about the interplay
have an effect on oxidative stress.
among oxidants, apoptosis, and aging will have to await
The answer to the question of whether or not the free
further experimentation (338a).
radical theory has yet been confirmed depends on ones
conception of what the theory predicts. In its broader
sense (oxidants contribute significantly to the process of D. Regulatory Roles for Oxidants and
degenerative senescence), the theory has clearly been Antioxidants: Signal Transduction
validated. In the more strict sense of the theory (oxidants
determine MLSP), the data are not yet conclusive, al- The discovery that NOj could serve as a signaling mol-
though a large body of consistent data has been generated. ecule has foreshadowed a more general recognition that
radical species are more than by-products. As mentioned
in section XVIIIB, the interaction between O20j and NOj
B. Nitric Oxide indicates that both may be regulatory molecules (58). In
other ways, the relationship between oxidants and signal-
A topic of great importance, which we have here ing is surfacing. For instance, the nitration of tyrosine resi-
largely omitted, is the role of nitric oxide (NOj) both as dues on a synthetic peptide substrate of cell cycle kinases
a damaging oxidant in its own right and as a signaling destroyed its ability to be phosphorylated, which illustrates
molecule that is susceptible to oxidative scavenging. NOj, a mechanism for an individual amino acid adduct to exert

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


572 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

an amplified effect (167). At a more physiological level, it Outstanding Investigator Grant CA39910 and National Institute
has been shown that the sensitivity of hippocampal brain of Environmental Health Sciences Grant ES-01896.
slices to muscarinic acetylcholine receptor agonists, which In writing a review on as broad a topic as the free radical
falls in an age-related fashion, can be restored by various theory, we have been forced to limit both the content and the
number of references cited. Although we have done our best to
antioxidant treatments and is potentiated (in old animals)
include recent work, omissions were inevitable. We apologize
by NOj-generating systems (155). to all authors whose work we have not managed to include and
Perhaps most intriguing have been discoveries, such direct readers to other recent reviews for material we have left
as the potential regulation of aconitases by O20j (95, 98) out.
and the realization that cell signaling via the Ras family Address for reprint requests: K. B. Beckman, Dept. of Mo-
of tyrosine kinases involves oxidants (147), that suggest lecular and Cell Biology, 401 Barker Hall, Univ. of California,
that oxidants play a role in signal transduction by design. Berkeley, CA 94720-3202.
These discoveries strengthen the free radical theory for
the following reason: if the regulated, enzymatic genera-
tion of oxidants plays a role in signaling pathways (legiti- REFERENCES
mate oxidants), then even a small increase in oxidant

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


generation by nonregulated pathways (illegitimate oxi- 1. ADELMAN, R., R. L. SAUL, AND B. N. AMES. Oxidative damage to
DNA: relation to species metabolic rate and life span. Proc. Natl.
dants) may have an exaggerated impact, by mimicking Acad. Sci. USA 85: 27062708, 1988.
the legitimate process. The arrival of redox regulation 2. AGARWAL, S., AND R. S. SOHAL. Relationship between aging and
as a viable field of inquiry, as evidenced by numerous susceptibility to protein oxidative damage. Biochem. Biophys. Res.
Commun. 194: 12031206, 1993.
topical recent reviews (66, 285, 321), may usher in a period 3. AGARWAL, S., AND R. S. SOHAL. Aging and protein oxidative dam-
when oxidants are seen as molecules of central physiolog- age. Mech. Ageing Dev. 75: 1119, 1994.
ical importance, in much the same way as are (for in- 4. ALBANES, D., O. P. HEINONEN, J. K. HUTTUNEN, P. R. TAYLOR,
J. VIRTAMO, B. K. EDWARDS, J. HAAPAKOSKI, M. RAUTALAHTI,
stance) lipid metabolites. As far as acceptance of the free A. M. HARTMAN, AND J. PALMGREN. Effects of alpha-tocopherol
radical theory is concerned, such a paradigm shift will and beta-carotene supplements on cancer incidence in the Alpha-
probably lead to de facto acceptance of its broader con- Tocopherol Beta-Carotene Cancer Prevention Study. Am. J. Clin.
Nutr. 62: 1427S-1430S, 1995.
ception (the idea that oxidants play a role in the process of 5. AMES, B. N., L. S. GOLD, AND W. C. WILLETT. The causes and
aging). After all, if it turns out that oxidants and oxidative prevention of cancer. Proc. Natl. Acad. Sci. USA 92: 52585265,
reactions are central to physiology, then how could they 1995.
6. AMES, B. N., M. K. SHIGENAGA, AND T. M. HAGEN. Oxidants, anti-
not be central to aging as well? If the free radical theory oxidants, and the degenerative diseases of aging. Proc. Natl. Acad.
once lacked appeal because oxidants appeared to be stoi- Sci. USA 90: 79157922, 1993.
chiometrically minor by-products, then those days may 7. ARNAIZ, S. L., M. TRAVACIO, S. LLESUY, AND A. BOVERIS. Hydro-
gen peroxide metabolism during peroxisome proliferation by feno-
soon be over. fibrate. Biochim. Biophys. Acta 1272: 175180, 1995.
On the other hand, in its more narrow conception 8. ARNHEIM, N., AND G. CORTOPASSI. Deleterious mitochondrial
(oxidants define MLSP), the free radical theory still DNA mutations accumulate in aging human tissues. Mutat. Res.
275: 157167, 1992.
lacks experimental confirmation. A hypothetical example 9. BAKER, G. T. D., AND R. L. SPROTT. Biomarkers of aging. Exp.
will illustrate this point. Even if mitogenic stimulation by Gerontol. 23: 223239, 1988.
Ras, for instance, were to involve oxidant generation in 10. BALIN, A. K., D. B. GOODMAN, H. RASMUSSEN, AND V. J. CRISTO-
FALO. The effect of oxygen and vitamin E on the life span of human
mice and humans, and the disregulation of Ras signaling diploid cells in vitro. J. Cell Biol. 74: 5867, 1977.
by illegitimate oxidants in aged individuals were shown 11. BANDY, B., AND A. J. DAVISON. Mitochondrial mutations may in-
to be important in degenerative senescence, the different crease oxidative stress: implications for carcinogenesis and aging?
Free Radical Biol. Med. 8: 523539, 1990.
life spans of these organisms still beg an explanation; it 12. BARJA, G., S. CADENAS, C. ROJAS, M. LOPEZ-TORRES, AND R.
still would remain to be explained why the process of PEREZ-CAMPO. A decrease of free radical production near critical
targets as a cause of maximum longevity in animals. Comp. Bio-
oxidative disregulation takes so much longer to occur in chem. Physiol. B Biochem. Mol. Biol. 108: 501512, 1994.
humans than in mice. In conclusion, despite the growing 13. BARJA, G., S. CADENAS, C. ROJAS, R. PEREZ-CAMPO, AND M.
concensus that the oxidants and oxygen free radicals are LOPEZ-TORRES. Low mitochondrial free radical production per
unit O2 consumption can explain the simultaneous presence of high
involved in degenerative senescence, countless mechanis- longevity and high aerobic metabolic rate in birds. Free Radical
tic questions remain to be uncovered, as well as a central Res. 21: 317327, 1994.
outstanding unknown: Do oxidants determine maximum 14. BARNETT, Y. A., AND C. M. KING. An investigation of antioxidant
status, DNA repair capacity and mutation as a function of age in
life span in humans and other organisms? humans. Mutat. Res. 338: 115128, 1995.
15. BEAL, M. F. Aging, energy, and oxidative stress in neurodegenera-
We thank Caleb Finch for critically reading the manuscript tive diseases. Ann. Neurol. 38: 357366, 1995.
and Irwin Fridovich for offering numerous detailed and useful 16. BECKMAN, K. B., AND B. N. AMES. Detection and quantification of
oxidative adducts of mitochondrial DNA. Methods Enzymol. 264:
comments. K. B. Beckman thanks Dr. I. Philipson for numerous 442453, 1996.
useful discussions. 17. BECKMAN, K. B., AND B. N. AMES. Oxidative decay of DNA. J.
B. N. Ames was supported by National Cancer Institute Biol. Chem. 272: 1963319636, 1997.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 573

18. BENZI, G., AND A. MORETTI. Are reactive oxygen species involved Oxidative DNA damage and senescence of human diploid fibroblast
in Alzheimers disease? Neurobiol. Aging. 16: 661674, 1995. cells. Proc. Natl. Acad. Sci. USA 92: 43374341, 1995.
19. BERLETT, B. S., R. L. LEVINE, AND E. R. STADTMAN. Comparison 42. CHIO, K. S., AND A. L. TAPPEL. Synthesis and characterization of
of the effects of ozone on the modification of amino acid residues the fluorescent products derived from malonaldehyde and amino
in glutamine synthetase and bovine serum albumin. J. Biol. Chem. acids. Biochemistry 8: 28212826, 1969.
271: 41774182, 1996. 43. CHOI, J. H., AND B. P. YU. Brain synaptosomal aging: free radicals
20. BEYER, W., J. IMLAY, AND I. FRIDOVICH. Superoxide dismutases. and membrane fluidity. Free Radical Biol. Med. 18: 133139, 1995.
Prog. Nucleic Acids Res. 40: 221253, 1991. 44. CHUNG, S. S., R. WEINDRUCH, S. R. SCHWARZE, D. I. MCKENZIE,
21. BITTLES, A. H. Evidence for and against the causal involvement AND J. M. AIKEN. Multiple age-associated mitochondrial DNA dele-
of mitochondrial DNA mutation in mammalian ageing. Mutat. Res. tions in skeletal muscle of mice. Aging 6: 193200, 1994.
275: 217225, 1992. 44a.COHEN, G., R. FAROOQUI, AND N. KESLER. Parkinson disease: a
22. BLOCK, G. Are clinical trials really the answer? Am. J. Clin. Nutr. new link between monoamine oxidase and mitochondrial electron
62, Suppl.: 1517515205, 1995. flow. Proc. Natl. Acad. Sci. USA 94: 48904894, 1997.
23. BOHR, V. A., AND R. M. ANSON. DNA damage, mutation and fine 44b.COLLINS, A., J. CADET, B. EPE, AND C. GEDIK. Problems in the
structure DNA repair in aging. Mutat. Res. 338: 2534, 1995. measurement of 8-oxoguanine in human DNA. Report of a work-
24. BOVERIS, A., AND E. CADENAS. Mitochondrial production of su- shop, DNA oxidation, held in Aberdeen, UK, 1921 January, 1997.
peroxide anions and its relationship to the antimycin insensitive Carcinogenesis 18: 18331836, 1997.
respiration. FEBS Lett. 54: 311314, 1975. 45. COMMONER, B., J. TOWNSEND, AND G. E. PAKE. Free radicals in
25. BOVERIS, A., E. CADENAS, AND A. O. STOPPANI. Role of ubiqui- biological materials. Nature 174: 689691, 1954.
none in the mitochondrial generation of hydrogen peroxide. Bio- 46. CONCONI, M., L. I. SZWEDA, R. L. LEVINE, E. R. STADTMAN, AND
chem. J. 156: 435444, 1976.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


B. FRIGUET. Age-related decline of rat liver multicatalytic protein-
26. BOVERIS, A., AND B. CHANCE. The mitochondrial generation of ase activity and protection from oxidative inactivation by heat-
hydrogen peroxide. General properties and effect of hyperbaric shock protein 90. Arch. Biochem. Biophys. 331: 232240, 1996.
oxygen. Biochem. J. 134: 707716, 1973. 47. CORRAL-DEBRINSKI, M., T. HORTON, M. T. LOTT, J. M. SHOFF-
27. BOVERIS, A., N. OSHINO, AND B. CHANCE. The cellular production NER, M. F. BEAL, AND D. C. WALLACE. Mitochondrial DNA dele-
of hydrogen peroxide. Biochem. J. 128: 617630, 1972. tions in human brain: regional variability and increase with ad-
28. BRUNK, U. T., C. B. JONES, AND R. S. SOHAL. A novel hypothesis vanced age. Nature Genet. 2: 324329, 1992.
of lipofuscinogenesis and cellular aging based on interactions be- 48. CORRAL-DEBRINSKI, M., T. HORTON, M. T. LOTT, J. M. SHOFF-
tween oxidative stress and autophagocytosis. Mutat. Res. 275: 395 NER, A. C. MCKEE, M. F. BEAL, B. H. GRAHAM, AND D. C. WAL-
403, 1992. LACE. Marked changes in mitochondrial DNA deletion levels in
29. BRUNK, U. T., H. ZHANG, H. DALEN, AND K. OLLINGER. Exposure Alzheimer brains. Genomics 23: 471476, 1994.
of cells to nonlethal concentrations of hydrogen peroxide induces 49. CORREA, P., AND V. W. CHEN. Gastric cancer. Cancer Surv. 20:
degeneration-repair mechanisms involving lysosomal destabiliza- 5576, 1994.
tion. Free Radical Biol. Med. 19: 813822, 1995. 50. CORTOPASSI, G., AND Y. LIU. Genotypic selection of mitochondrial
29a.CADET, J., T. DOUKI, AND J.-L. RAVANAT. Artifacts associated and oncogenic mutations in human tissue suggests mechanisms of
with the measurement of oxidized DNA bases. Environ. Health age-related pathophysiology. Mutat. Res. 338: 151159, 1995.
Perspect. 105: 10341039, 1997. 51. CORTOPASSI, G., Y. LIU, AND T. HUTCHIN. Degeneration of human
30. CAMPISI, J. Replicative senescence: an old lives tale? Cell 84: 497 oncogenes and mitochondrial genes occurs in cells that exhibit
500, 1996. age-related pathology. Exp. Gerontol. 31: 253265, 1996.
31. CAO, G., AND R. G. CUTLER. Protein oxidation and aging. I. Diffi- 52. CORTOPASSI, G., AND E. WANG. Modelling the effects of age-
culties in measuring reactive protein carbonyls in tissues using related mtDNA mutation accumulation: complex I deficiency, su-
2,4-dinitrophenylhydrazine. Arch. Biochem. Biophys. 320: 106114, peroxide and cell death. Biochim. Biophys. Acta 1271: 171176,
1995.
1995.
32. CAO, G., AND R. G. CUTLER. Protein oxidation and aging. II. Diffi-
52a.CORTOPASSI, G., AND E. WANG. There is substantial agreement
culties in measuring alkaline protease activity in tissues using the
among interspecies estimates of DNA repair activity. Mech. Ageing
fluorescamine procedure. Arch. Biochem. Biophys. 320: 195201,
Dev. 91: 211218, 1996.
1995.
53. CORTOPASSI, G. A., AND N. ARNHEIM. Detection of a specific
33. CARNEY, J. M., C. D. SMITH, A. M. CARNEY, AND D. A. BUT-
mitochondrial DNA deletion in tissues of older humans. Nucleic
TERFIELD. Aging- and oxygen-induced modifications in brain bio-
chemistry and behavior. Ann. NY Acad. Sci. 738: 4453, 1994. Acids Res. 18: 69276933, 1990.
34. CARNEY, J. M., P. E. STARKE-REED, C. N. OLIVER, R. W. LAN- 54. CORTOPASSI, G. A., D. SHIBATA, N. W. SOONG, AND N. ARN-
DUM, M. S. CHENG, J. F. WU, AND R. A. FLOYD. Reversal of age- HEIM. A pattern of accumulation of a somatic deletion of mitochon-
related increase in brain protein oxidation, decrease in enzyme drial DNA in aging human tissues. Proc. Natl. Acad. Sci. USA 89:
activity, and loss in temporal and spatial memory by chronic admin- 73707374, 1992.
istration of the spin-trapping compound N-tert-butyl-alpha-phenyl- 55. CRISTOFALO, V. J. Ten years later: what have we learned about
nitrone. Proc. Natl. Acad. Sci. USA 88: 36333636, 1991. human aging from studies of cell cultures? Gerontologist 36: 737
35. CERAMI, A. Hypothesis. Glucose as a mediator of aging. J. Am. 741, 1996.
Geriatr. Soc. 33: 626634, 1985. 56. CRISTOFALO, V. J., AND R. J. PIGNOLO. Replicative senescence of
36. CHAN, P. H., C. J. EPSTEIN, Y. LI, T. T. HUANG, E. CARLSON, human fibroblast-like cells in culture. Physiol. Rev. 73: 617638,
H. KINOUCHI, G. YANG, H. KAMII, S. MIKAWA, AND T. KONDO. 1993.
Transgenic mice and knockout mutants in the study of oxidative 56a.CROTEAU, D. L., AND V. A. BOHR. Repair of oxidative damage to
stress in brain injury. J. Neurotrauma 12: 815824, 1995. nuclear and mitochondrial DNA in mammalian cells. J. Biol. Chem.
37. CHANCE, B., H. SIES, AND A. BOVERIS. Hydroperoxide metabo- 272: 2540925412, 1997.
lism in mammalian organs. Physiol. Rev. 59: 527605, 1979. 56b.CUNNINGHAM, R. P. DNA repair: caretakers of the genome? Curr.
38. CHANOCK, S. J., B. J. EL, R. M. SMITH, AND B. M. BABIOR. The Biol. 7: R576R579, 1997.
respiratory burst oxidase. J. Biol. Chem. 269: 2451924522, 1994. 57. CUTLER, R. G. Antioxidants and aging. Am. J. Clin. Nutr. 53,
39. CHEN, J. J., AND B. P. YU. Alterations in mitochondrial membrane Suppl.: 373S379S, 1991.
fluidity by lipid peroxidation products. Free Radical Biol. Med. 17: 58. DARLEY-USMAR, V., H. WISEMAN, AND B. HALLIWELL. Nitric ox-
411418, 1994. ide and oxygen radicals: a question of balance. FEBS Lett. 369:
40. CHEN, Q., AND B. N. AMES. Senescence-like growth arrest induced 131135, 1995.
by hydrogen peroxide in human diploid fibroblast F65 cells. Proc. 59. DARR, D., AND I. FRIDOVICH. Adaptation to oxidative stress in
Natl. Acad. Sci. USA 91: 41304134, 1994. young, but not in mature or old, Caenorhabditis elegans. Free
41. CHEN, Q., A. FISCHER, J. D. REAGAN, L. J. YAN, AND B. N. AMES. Radical Biol. Med. 18: 195201, 1995.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


574 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

60. DAVIES, K. J. Oxidative stress: the paradox of aerobic life. Bio- mutations in the brain: relationship to aging? Mutat. Res. 338: 173
chem. Soc. Symp. 61: 131, 1995. 182, 1995.
61. DAVIES, M. J., S. FU, AND R. T. DEAN. Protein hydroperoxides can 79a.EWBANK, J. J., T. M. BARNES, B. LAKOWSKI, M. LUSSIER, H.
give rise to reactive free radicals. Biochem. J. 305: 643649, 1995. BUSSEY, AND S. HEKIMI. Structural and functional conservation
62. DAVIS, R. E. M. S., C. HERRNSTADT, S. S. GHOSH, E. FAHY, L. A. of the Caenorhabditis elegans timing gene clk-1 (see comments).
SHINOBU, D. GALASKO, L. J. THAL, M. F. BEAL, N. HOWELL, AND Science 275: 980983, 1997.
W. D. PARKER. Mutations in mitochondrial cytochrome c oxidase 80. FAHN, S., AND G. COHEN. The oxidant stress hypothesis in Parkin-
genes segregate with late-onset Alzheimers disease. Proc. Natl. sons disease: evidence supporting it. Ann. Neurol. 32: 804812,
Acad. Sci. USA 94: 45264531, 1997. 1992.
63. DEAN, R. T., J. GEBICKI, S. GIESEG, A. J. GRANT, AND J. A. SIMP- 81. FEIG, D. I., T. M. REID, AND L. A. LOEB. Reactive oxygen species
SON. Hypothesis: a damaging role in aging for reactive protein in tumorigenesis. Cancer Res. 54, Suppl.: 1890s1894s, 1994.
oxidation products? Mutat. Res. 275: 387393, 1992. 82. FEUERS, R. J., R. WEINDRUCH, AND R. W. HART. Caloric restric-
64. DEAN, R. T., S. GIESEG, AND M. J. DAVIES. Reactive species and tion, aging, and antioxidant enzymes. Mutat. Res. 295: 191200,
their accumulation on radical-damaged proteins. Trends Biochem. 1993.
Sci. 18: 437441, 1993. 83. FINCH, C. E. Longevity, Senescence, and The Genome. Chicago,
65. DE LA PAZ, M. A., J. ZHANG, AND I. FRIDOVICH. Antioxidant en- IL: Univ. of Chicago Press, 1990.
zymes of the human retina: effect of age on enzyme activity of 84. FLEMING, J. E., J. MIQUEL, S. F. COTTRELL, L. S. YENGOYAN,
macula and periphery. Curr. Eye Res. 15: 273278, 1996. AND A. C. ECONOMOS. Is cell aging caused by respiration-depen-
66. DEMPLE, B. Redox signaling and gene control in the Escherichia dent injury to the mitochondrial genome? Gerontology 28: 4453,
coli soxRS oxidative stress regulona review. Gene 179: 5357, 1982.
85. FLEMING, J. E., I. REVEILLAUD, AND A. NIEDZWIECKI. Role of

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


1996.
67. DENNOG, C., A. HARTMANN, G. FREY, AND G. SPEIT. Detection oxidative stress in Drosophila aging. Mutat. Res. 275: 267279,
of DNA damage after hyperbaric oxygen (HBO) therapy. Mutagene- 1992.
sis 11: 605609, 1996. 86. FLINT, D. H., J. F. TUMINELLO, AND M. H. EMPTAGE. The inacti-
68. DIMRI, G. P., X. LEE, G. BASILE, M. ACOSTA, G. SCOTT, C. vation of Fe-S cluster containing hydrolyases by superoxide. J.
ROSKELLEY, E. E. MEDRANO, M. LINSKENS, I. RUBELJ, O. PER- Biol. Chem. 268: 2236922376, 1993.
EIRA-SMITH, M. PEACOCKE, AND J. CAMPISI. A biomarker that 87. FLOYD, R. A., AND J. M. CARNEY. Free radical damage to protein
identifies senescent human cells in culture and in aging skin in and DNA: mechanisms involved and relevant observations on brain
vivo. Proc. Natl. Acad. Sci. USA 92: 93639367, 1995. undergoing oxidative stress. Ann. Neurol. 32, Suppl.: S22S27,
69. DIONISI, O., T. GALEOTTI, T. TERRANOVA, AND A. AZZI. Superox- 1992.
ide radicals and hydrogen peroxide formation in mitochondria from 88. FORMAN, H. J., AND A. AZZI. On the virtual existence of superoxide
normal and neoplastic tissues. Biochim. Biophys. Acta 403: 292 anion in mitochondria: thoughts regarding its role in pathophysiol-
300, 1975. ogy. FASEB J. 11: 374375, 1997.
70. DIZDAROGLU, M. Measurement of radiation-induced damage to 89. FORMAN, H. J., AND J. A. KENNEDY. Role of superoxide radical
DNA at the molecular level. Int. J. Radiat. Biol. 61: 175183, 1992. in mitochondrial dehydrogenase reactions. Biochem. Biophys. Res.
71. DIZDAROGLU, M. Oxidative damage to DNA in mammalian chro- Commun. 60: 10441050, 1974.
matin. Mutat. Res. 275: 331342, 1992. 90. FORSTER, M. J., A. DUBEY, K. M. DAWSON, W. A. STUTTS, H.
71a.DOLLE, M. E., H. GIESE, C. L. HOPKINS, H. J. MARTUS, J. M. LAL, AND R. S. SOHAL. Age-related losses of cognitive function and
HAUSDORFF, AND J. VIJG. Rapid accumulation of genome rear- motor skills in mice are associated with oxidative protein damage
rangements in liver but not in brain of old mice. Nature Genet. 17: in the brain. Proc. Natl. Acad. Sci. USA 93: 47654769, 1996.
431434, 1997. 91. FRIDOVICH, I. Superoxide radical and superoxide dismutases.
Annu. Rev. Biochem. 64: 97112, 1995.
72. DORMAN, J. B., B. ALBINDER, T. SHROYER, AND C. KENYON.
92. FRIGUET, B., E. R. STADTMAN, AND L. I. SZWEDA. Modification
The age-1 and daf-2 genes function in a common pathway to control
of glucose-6-phosphate dehydrogenase by 4-hydroxy-2-nonenal.
the life span of Caenorhabditis elegans. Genetics 141: 13991406,
Formation of cross-linked protein that inhibits the multicatalytic
1995.
protease. J. Biol. Chem. 269: 2163921643, 1994.
73. DUBEY, A., M. J. FORSTER, AND R. S. SOHAL. Effect of the spin-
93. FRIGUET, B., L. I. SZWEDA, AND E. R. STADTMAN. Susceptibility
trapping compound N-tert-butyl-alpha-phenylnitrone on protein ox-
of glucose-6-phosphate dehydrogenase modified by 4-hydroxy-2-
idation and life span. Arch. Biochem. Biophys. 324: 249254, 1995. nonenal and metal-catalyzed oxidation to proteolysis by the multi-
74. DUPRAT, F., E. GUILLEMARE, G. ROMEY, M. FINK, F. LESAGE, catalytic protease. Arch. Biochem. Biophys. 311: 168173, 1994.
M. LAZDUNSKI, AND E. HONORE. Susceptibility of cloned K/ chan- 93a.GABBITA, S. P., D. A. BUTTERFIELD, K. HENSLEY, W. SHAW, AND
nels to reactive oxygen species. Proc. Natl. Acad. Sci. USA 92: J. M. CARNEY. Aging and caloric restriction affect mitochondrial
1179611800, 1995. respiration and lipid membrane status: an electron paramagnetic
74a.DYPBUKT, J. M., M. ANKARCRONA, M. BURKITT, A. SJOHOLM, resonance investigation. Free Radical Biol. Med. 23: 191201, 1997.
K. STROM, S. ORRENIUS, AND P. NICOTERA. Different prooxidant 94. GAFNI, A. Altered protein metabolism in aging. Annu. Rev. Geron-
levels stimulate growth, trigger apoptosis, or produce necrosis of tol. Geriatr. 10: 117131, 1990.
insulin-secreting RINm5F cells. The role of intracellular poly- 95. GARDNER, P. R., AND I. FRIDOVICH. Superoxide sensitivity of the
amines. J. Biol. Chem. 269: 3055330560, 1994. Escherichia coli aconitase. J. Biol. Chem. 266: 1932819333, 1991.
75. EDAMATSU, R., A. MORI, AND L. PACKER. The spin-trap N-tert- 96. GARDNER, P. R., AND I. FRIDOVICH. Inactivation-reactivation of
alpha-phenyl-butylnitrone prolongs the life span of the senescence aconitase in Escherichia coli. A sensitive measure of superoxide
accelerated mouse. Biochem. Biophys. Res. Commun. 211: 847 radical. J. Biol. Chem. 267: 87578763, 1992.
849, 1995. 97. GARDNER, P. R., D. D. NGUYEN, AND C. W. WHITE. Aconitase
76. EDRIS, W., B. BURGETT, O. C. STINE, AND C. R. FILBURN. Detec- is a sensitive and critical target of oxygen poisoning in cultured
tion and quantitation by competitive PCR of an age-associated in- mammalian cells and in rat lungs. Proc. Natl. Acad. Sci. USA 91:
crease in a 4.8-kb deletion in rat mitochondrial DNA. Mutat. Res. 1224812252, 1994.
316: 6978, 1994. 98. GARDNER, P. R., I. RAINERI, L. B. EPSTEIN, AND C. W. WHITE.
77. ELGAWISH, A., M. GLOMB, M. FRIEDLANDER, AND V. M. MON- Superoxide radical and iron modulate aconitase activity in mamma-
NIER. Involvement of hydrogen peroxide in collagen cross-linking lian cells. J. Biol. Chem. 270: 1339913405, 1995.
by high glucose in vitro and in vivo. J. Biol. Chem. 271: 12964 99. GAZIEV, A. I., A. PODLUTSKY, B. M. PANFILOV, AND R. BRAD-
12971, 1996. BURY. Dietary supplements of antioxidants reduce hprt mutant
78. EPE, B. DNA damage profiles induced by oxidizing agents. Rev. frequency in splenocytes of aging mice. Mutat. Res. 338: 7786,
Physiol. Biochem. Pharmacol. 127: 223249, 1996. 1995.
79. EVANS, D. A., J. P. BURBACH, AND F. W. VAN LEEUWEN. Somatic 100. GAZIEV, A. I., G. R. SOLOGUB, L. A. FOMENKO, S. I. ZAICHKINA,

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 575

N. I. KOSYAKOVA, AND R. J. BRADBURY. Effect of vitamin-antioxi- 124. HARMAN, D. Free radical involvement in aging. Pathophysiology
dant micronutrients on the frequency of spontaneous and in vitro and therapeutic implications. Drugs Aging 3: 6080, 1993.
gamma-ray-induced micronuclei in lymphocytes of donors: the age 125. HARMAN, D. Free-radical theory of aging. Increasing the functional
factor. Carcinogenesis 17: 493499, 1996. life span. Ann. NY Acad. Sci. 717: 115, 1994.
101. GEBICKI, J. M. Protein hydroperoxides as new reactive oxygen 126. HARMAN, D. Aging and disease: extending functional life span.
species. Redox Report 3: 99110, 1997. Ann. NY Acad. Sci. 786: 321336, 1996.
102. GIESEG, S. P., J. A. SIMPSON, T. S. CHARLTON, M. W. DUNCAN, 127. HARMAN, D. H. Free radical theory of aging: effects of antioxidants
AND R. T. DEAN. Protein-bound 3,4-dihydroxyphenylalanine is a on mitochondrial function. Age 10: 5861, 1987.
major reductant formed during hydroxyl radical damage to pro- 127a.HART, R. W., AND R. B. SETLOW. Correlation between deoxyribo-
teins. Biochemistry 32: 47804786, 1993. nucleic acid excision-repair and life-span in a number of mamma-
103. GILCHREST, B. A., AND V. A. BOHR. Aging processes, DNA dam- lian species. Proc. Natl. Acad. Sci. USA 71: 21692173, 1974.
age, and repair. FASEB J. 11: 322330, 1997. 127b.HEKIMI, S., B. LAKOWSKI, T. M. BARNES, AND J. J. EWBANK.
104. GIULIVI, C., R. E. PACIFICI, AND K. J. DAVIES. Exposure of hy- Molecular genetics of life span in C. elegans: how much does it
drophobic moieties promotes the selective degradation of hydrogen teach us? Trends Genet. 14: 1420, 1998.
peroxide-modified hemoglobin by the multicatalytic proteinase 127c.HELBOCK, H. J., K. B. BECKMAN, M. K. SHIGENAGA, P. W. WAL-
complex, proteasome. Arch. Biochem. Biophys. 311: 329341, TER, A. A. WOODALL, H. C. YEO, AND B. N. AMES. DNA oxidation
1994. matters: the HPLC-electrochemical detection assay of 8-oxo-deoxy-
105. GOEPTAR, A. R., H. SCHEERENS, AND N. P. VERMEULEN. Oxygen guanosine and 8-oxo-guanine. Proc. Natl. Acad. Sci. USA 95: 288
and xenobiotic reductase activities of cytochrome P450. Crit. Rev. 293, 1998.
Toxicol. 25: 2565, 1995. 128. HEMPELMANN, L. H., AND J. G. HOFFMAN. Practical aspects of

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


106. GOTO, S., AND A. NAKAMURA. Age-associated, oxidatively modi- radiation injury. Annu. Rev. Nucl. Sci. 3: 369389, 1953.
fied proteins: a critical evaluation. Age 20: 8189, 1997. 129. HENNEKENS, C. H., J. E. BURING, J. E. MANSON, M. STAMPFER,
106a.GRAY, M. D., J. C. SHEN, A. S. KAMATH-LOEB, A. BLANK, B. L. B. ROSNER, N. R. COOK, C. BELANGER, F. LAMOTTE, J. M. GAZI-
SOPHER, G. M. MARTIN, J. OSHIMA, AND L. A. LOEB. The Werner ANO, P. M. RIDKER, W. WILLETT, AND R. PETO. Lack of effect of
syndrome protein is a DNA helicase. Nature Genet. 17: 100103, long-term supplementation with beta carotene on the incidence of
1997. malignant neoplasms and cardiovascular disease. N. Engl. J. Med.
107. GREIDER, C. W. Mammalian telomere dynamics: healing, fragmen- 334: 11451149, 1996.
tation shortening and stabilization. Curr. Opin. Genet. Dev. 4: 203 130. HENSLEY, K., D. A. BUTTERFIELD, N. HALL, P. COLE, R. SUBRA-
211, 1994. MANIAM, R. MARK, M. P. MATTSON, W. R. MARKESBERY, M. E.
108. GROLLMAN, A. P., AND M. MORIYA. Mutagenesis by 8-oxoguanine: HARRIS, M. AKSENOV, M. AKSENOVA, J. F. WU, AND J. M. CAR-
an enemy within. Trends Genet. 9: 246249, 1993. NEY. Reactive oxygen species as causal agents in the neurotoxicity
109. GUTTERIDGE, J. M., AND B. HALLIWELL. The measurement and of the Alzheimers disease-associated amyloid beta peptide. Ann.
mechanism of lipid peroxidation in biological systems. Trends Bio- NY Acad. Sci. 786: 120134, 1996.
chem. Sci. 15: 129135, 1990. 130a.HERRERO, A., AND G. BARJA. Sites and mechanisms responsible
110. HAGEN, T. M., D. L. YOWE, J. C. BARTHOLOMEW, C. M. WEHR, for the low rate of free radical production of heart mitochondria
K. L. DO, J. Y. PARK, AND B. N. AMES. Mitochondrial decay in hepa- in the long-lived pigeon. Mech. Ageing Dev. 98: 95111, 1997.
tocytes from old rats: membrane potential declines, heterogeneity 131. HIGAMI, Y., I. SHIMOKAWA, T. OKIMOTO, AND T. IKEDA. An age-
and oxidants increase. Proc. Natl. Acad. Sci. USA 94: 30643069, related increase in the basal level of DNA damage and DNA vulnera-
1997. bility to oxygen radicals in the individual hepatocytes of male F344
111. HALEY, Z. V., AND A. RICHARDSON. Effect of dietary restriction rats. Mutat. Res. 316: 5967, 1994.
on DNA repair and DNA damage. Mutat. Res. 295: 237245, 1993. 131a.HIGUCHI, Y., AND S. LINN. Purification of all forms of HeLa cell
112. HALLIWELL, B. Free radicals, reactive oxygen species, and human mitochondrial DNA and assessment of damage to it caused by
disease: a critical evaluation with special reference to atherosclero- hydrogen peroxide treatment of mitochondria or cells. J. Biol.
sis. Br. J. Exp. Pathol. 70: 737757, 1989. Chem. 270: 79507956, 1995.
113. HALLIWELL, B., AND M. DIZDAROGLU. The measurement of oxida- 131b.HIRANO, M., A. SHTILBANS, R. MAYEUX, M. M. DAVIDSON, S.
tive damage to DNA by HPLC and GC/MS techniques. Free Radical DIMAURO, J. A. KNOWLES, AND E. A. SCHON. Apparent mtDNA
Res. Commun. 16: 7587, 1992. heteroplasmy in Alzheimers disease patients and in normals due
114. HALLIWELL, B., AND J. M. GUTTERIDGE. Oxygen free radicals to PCR amplification of nucleus-embedded mtDNA pseudogenes.
and iron in relation to biology and medicine: some problems and Proc. Natl. Acad. Sci. USA 94: 1489414899. 1997.
concepts. Arch. Biochem. Biophys. 246: 501514, 1986. 132. HIRANO, T., R. YAMAGUCHI, S. ASAMI, N. IWAMOTO, AND H.
115. HALLIWELL, B. H., AND J. M. C. GUTTERIDGE. Free Radicals in KASAI. 8-Hydroxyguanine levels in nuclear DNA and its repair ac-
Biology and Medicine. Oxford, UK: Oxford Univ. Press, 1989. tivity in rat organs associated with age. J. Gerontol. 51: 303307,
116. HANAWALT, P. C. DNA repair comes of age. Mutat. Res. 336: 101 1996.
113, 1995. 133. HIRANO, T., Y. YAMAGUCHI, H. HIRANO, AND H. KASAI. Age-
117. HANAWALT, P. C., P. GEE, L. HO, R. K. HSU, AND C. J. KANE. associated change of 8-hydroxyguanine repair activity in cultured
Genomic heterogeneity of DNA repair. Role in aging? Ann. NY human fibroblasts. Biochem. Biophys. Res. Commun. 214: 1157
Acad. Sci. 663: 1725, 1992. 1162, 1995.
118. HANSFORD, R. G., B. A. HOGUE, AND V. MILDAZIENE. Depen- 134. HO, Y.-S., J.-L. MAGNENAT, R. T. BRONSON, J. CAO, M. GAR-
dence of H2O2 formation by rat heart mitochondria on substrate GANO, M. SUGAWARA, AND C. D. FUNK. Mice deficient in cellular
availablity and donor age. J. Bioenerg. Biomembr. 29: 8995, 1997. glutathione peroxidase develop normally and show no increased
119. HARDMEIER, R., H. HOEGER, S. FANG-KIRCHER, A. KHOSCHS- sensitivity to hyperoxia. J. Biol. Chem. 272: 1664416651, 1997.
ORUR, AND G. LUBEC. Transcription and activity of antioxidant 135. HOCHSCHILD, R. Lysosomes, membranes and aging. Exp. Geron-
enzymes after ionizing irradiation in radiation-resistant and radia- tol. 6: 153166, 1971.
tion-sensitive mice. Proc. Natl. Acad. Sci. USA 94: 75727576, 1997. 136. HOLLIDAY, R. Mutations and epimutations in mammalian cells.
120. HARMAN, D. Aging: a theory based on free radical and radiation Mutat. Res. 250: 351363, 1991.
chemistry. J. Gerontol. 2: 298300, 1956. 137. HOLMES, G. E., C. BERNSTEIN, AND H. BERNSTEIN. Oxidative
121. HARMAN, D. The biologic clock: the mitochondria? J. Am. Geriatr. and other DNA damages as the basis of aging: a review. Mutat.
Soc. 20: 145147, 1972. Res. 275: 305315, 1992.
122. HARMAN, D. The aging process. Proc. Natl. Acad. Sci. USA 78: 138. HONDA, S., N. ISHII, K. SUZUKI, AND M. MATSUO. Oxygen-depen-
71247128, 1981. dent perturbation of life span and aging rate in the nematode. J.
123. HARMAN, D. Free radical theory of aging. Mutat. Res. 275: 257 Gerontol. 48: 5761, 1993.
266, 1992. 139. HONDA, S., AND M. MATSUO. The sensitivity to hyperbaric oxygen

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


576 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

of human diploid fibroblasts during ageing in vitro. Mech. Ageing 2*-deoxyguanosine, as a marker of cellular oxidative stress during
Dev. 12: 3137, 1980. carcinogenesis. Mutat. Res. 387: 147163, 1997.
140. HONDA, S., AND M. MATSUO. Shortening of the in vitro life span 157. KASAI, H., Y. OKADA, S. NISHIMURA, M. S. RAO, AND J. K. REDDY.
of human diploid fibroblasts exposed to hyperbaric oxygen. Exp. Formation of 8-hydroxydeoxyguanosine in liver DNA of rats follow-
Gerontol. 18: 339345, 1983. ing long-term exposure to a peroxisome proliferator. Cancer Res.
141. HONDA, S., AND M. MATSUO. Lack of recovery from oxygen-in- 49: 26032605, 1989.
duced damage to colony formation and DNA synthesis in senescent 158. KENYON, C. Ponce delegans: genetic quest for the fountain of
human diploid fibroblasts. Mech. Ageing Dev. 40: 8187, 1987. youth. Cell 84: 501504, 1996.
142. HORI, O., J. BRETT, T. SLATTERY, R. CAO, J. ZHANG, J. X. CHEN, 159. KEYER, K., A. S. GORT, AND J. A. IMLAY. Superoxide and the pro-
M. NAGASHIMA, E. R. LUNDH, S. VIJAY, D. NITECKI, AND A. L. duction of oxidative DNA damage. J. Bacteriol. 177: 67826790,
ET. The receptor for advanced glycation end products (RAGE) is a 1995.
cellular binding site for amphoterin. Mediation of neurite outgrowth 160. KIKUGAWA, K., T. KATO, AND A. HAYASAKA. Formation of dityro-
and co-expression of rage and amphoterin in the developing ner- sine and other fluorescent amino acids by reaction of amino acids
vous system. J. Biol. Chem. 270: 2575225761, 1995. with lipid hydroperoxides. Lipids 26: 922929, 1991.
143. HORTON, T. M., B. H. GRAHAM, D. M. CORRAL, J. M. SHOFFNER, 160a.KIMURA, K. D., H. A. TISSENBAUM, Y. LIU, AND G. RUVKUN.
A. E. KAUFMAN, M. F. BEAL, AND D. C. WALLACE. Marked in- Daf-2, an insulin receptor-like gene that regulates longevity and
crease in mitochondrial DNA deletion levels in the cerebral cortex diapause in Caenorhabditis elegans. Science 277: 942946, 1997.
of Huntingtons disease patients. Neurology 45: 18791883, 1995. 161. KING, C. M., AND Y. A. BARNETT. Oxidative stress and human age-
144. HOSOKAWA, H., N. ISHII, H. ISHIDA, K. ICHIMORI, H. NAKA- ing (Abstract). Biochem. Soc. Trans. 23: 375S, 1995.
ZAWA, AND K. SUZUKI. Rapid accumulation of fluorescent material 161a.KING, C. M. H. E. BRISTOW-CRAIG, E. S. GILLESPIE, AND Y. A.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


with aging in an oxygen-sensitive mutant mev-1 of Caenorhabditis BARNETT. In vivo antioxidant status, DNA damage, mutation and
elegans. Mech. Ageing Dev. 74: 161170, 1994. DNA repair capacity in cultured lymphocytes from healthy 75- to
145. HU, J. J., N. DUBIN, D. KURLAND, B.-L. MA, AND G. C. ROUSH. 80-year-old humans. Mutat. Res. 377: 137147, 1997.
The effects of hydrogen peroxide on DNA repair activities. Mutat. 162. KIRKWOOD, T. B. Comparative life spans of species: why do spe-
Res. 336: 193201, 1995. cies have the life spans they do? Am. J. Clin. Nutr. 55, Suppl.:
145a.HUNOT, S., B. BRUGG, D. RICARD, P. P. MICHEL, M. P. MURIEL, 1191S1195S, 1992.
M. RUBERG, B. A. FAUCHEUX, Y. AGID, AND E. C. HIRSCH. Nu- 163. KIRKWOOD, T. B., AND M. R. ROSE. Evolution of senescence: late
clear translocation of NF-kappaB is increased in dopaminergic neu- survival sacrificed for reproduction. Philos. Trans. R. Soc. Lond B
rons of patients with Parkinson disease. Proc. Natl. Acad. Sci. USA Biol. Sci. 332: 1524, 1991.
94: 75317536, 1997. 164. KIRKWOOD, T. B. L. Evolution and aging. Nature 270: 301304,
146. IMLAY, J. A., AND I. FRIDOVICH. Assay of metabolic superoxide 1977.
production in Escherichia coli. J. Biol. Chem. 266: 69576965, 165. KNEEPKENS, C. M., G. LEPAGE, AND C. C. ROY. The potential of
1991. the hydrocarbon breath test as a measure of lipid peroxidation.
147. IRANI, K., Y. XIA, J. L. ZWEIER, S. J. SOLLOTT, C. J. DER, E. R. Free Radical Biol. Med. 17: 127160, 1994.
FEARON, M. SUNDARESAN, T. FINKEL, AND P. J. GOLD- 166. KNIGHT, J. A. The process and theories of aging. Ann. Clin. Lab.
SCHMIDT-CLERMONT. Mitogenic signaling mediated by oxidants Sci. 25: 112, 1995.
in ras-transformed fibroblasts. Science 275: 16491652, 1997. 167. KONG, S. K., M. B. YIM, E. R. STADTMAN, AND P. B. CHOCK. Per-
148. ISHII, N., K. TAKAHASHI, S. TOMITA, T. KEINO, S. HONDA, K. oxynitrite disables the tyrosine phosphorylation regulatory mecha-
YOSHINO, AND K. SUZUKI. A methyl viologen-sensitive mutant of nism: lymphocyte-specific tyrosine kinase fails to phosphorylate
the nematode Caenorhabditis elegans. Mutat. Res. 237: 165171, nitrated cdc2(620)NH2 peptide. Proc. Natl. Acad. Sci. USA 93:
1990. 33773382, 1996.
149. IVY, G. O., F. SCHOTTLER, J. WENZEL, M. BAUDRY, AND G. 168. KOOP, D. R. Oxidative and reductive metabolism by cytochrome
LYNCH. Inhibitors of lysosomal enzymes: accumulation of lipofus- P450 2E1. FASEB J. 6: 724730, 1992.
cin-like dense bodies in the brain. Science 226: 985987, 1984. 169. KOSTER, J. F., AND W. SLUITER. Is increased tissue ferritin a risk
149a.JACOBSON, M. D. Reactive oxygen species and programmed cell factor for atherosclerosis and ischaemic heart disease? Br. Heart
death. Trends Biochem. Sci. 21: 8386, 1996. J. 73: 208209, 1995.
150. JI, L. L., D. DILLON, AND E. WU. Alteration of antioxidant enzymes 170. KOWALD, A., AND T. B. KIRKWOOD. A network theory of ageing:
with aging in rat skeletal muscle and liver. Am. J. Physiol. 258 the interactions of defective mitochondria, aberrant proteins, free
(Regulatory Integrative Comp. Physiol. 27): R918R923, 1990. radicals and scavengers in the ageing process. Mutat. Res. 316:
151. JI, L. L., D. DILLON, AND E. WU. Myocardial aging: antioxidant 209236, 1996.
enzyme systems and related biochemical properties. Am. J. Phys- 171. KRUK, P. A., N. J. RAMPINO, AND V. A. BOHR. DNA damage and
iol. 260 (Regulatory Integrative Comp. Physiol. 29): R386R392, repair in telomeres: relation to aging. Proc. Natl. Acad. Sci. USA
1991. 92: 258262, 1995.
152. JOHNSON, T. E., AND G. J. LITHGOW. The search for the genetic 172. KU, H. H., U. T. BRUNK, AND R. S. SOHAL. Relationship between
basis of aging: the identification of gerontogenes in the nematode mitochondrial superoxide and hydrogen peroxide production and
Caenorhabditis elegans. J. Am. Geriatr. Soc. 40: 936945, 1992. longevity of mammalian species. Free Radical Biol. Med. 15: 621
153. JOHNSON, T. E., P. M. TEDESCO, AND G. J. LITHGOW. Comparing 627, 1993.
mutants, selective breeding, and transgenics in the dissection of 173. KU, H. H., AND R. S. SOHAL. Comparison of mitochondrial pro-
aging processes of Caenorhabditis elegans. Genetica 91: 6577, oxidant generation and anti-oxidant defenses between rat and pi-
1993. geon: possible basis of variation in longevity and metabolic poten-
154. JOHNSON, T. M., Z. X. YU, V. J. FERRANS, R. A. LOWENSTEIN, tial. Mech. Ageing Dev. 72: 6776, 1993.
AND T. FINKEL. Reactive oxygen species are downstream media- 174. KUMARI, M. V., T. YONEDA, AND M. HIRAMATSU. Scavenging ac-
tors of p53-dependent apoptosis. Proc. Natl. Acad. Sci. USA 93: tivity of beta catechin on reactive oxygen species generated by
1184811852, 1996. photosensitization of riboflavin. Biochem. Mol. Biol. Int. 38: 1163
155. JOSEPH, J. A., R. VILLALOBOS-MOLINA, N. DENISOVA, S. ERAT, 1170, 1996.
N. JIMENEZ, AND J. STRAIN. Increased sensitivity to oxidative 175. KUMARI, M. V., T. YONEDA, AND M. HIRAMATSU. Effect of beta
stress and the loss of muscarinic receptor responsiveness in senes- CATECHIN on the life span of senescence accelerated mice (SAM-
cence. Ann. NY Acad. Sci. 786: 112119, 1996. P8 strain). Biochem. Mol. Biol. Int. 41: 10051011, 1997.
156. KANEKO, T., S. TAHARA, AND M. MATSUO. Nonlinear accumula- 176. KUO, C. F., T. MASHINO, AND I. FRIDOVICH. Alpha,beta-
tion of 8-hydroxy-2*-deoxyguanosine, a marker of oxidized DNA dihydroxyisovalerate dehydratase. A superoxide-sensitive enzyme.
damage, during aging. Mutat. Res. 316: 277285, 1996. J. Biol. Chem. 262: 47244727, 1987.
156a.KASAI, H. Analysis of a form of oxidative DNA damage, 8-hydroxy- 177. LAKE, B. G. Mechanisms of hepatocarcinogenicity of peroxisome-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 577

proliferating drugs and chemicals. Annu. Rev. Pharmacol. Toxicol. compounds are mutagens in Salmonella tester strain TA104. Mutat.
35: 483507, 1995. Res. 148: 2534, 1985.
178. LARSEN, P. L. Aging and resistance to oxidative damage in Caeno- 195. MARTIN, G. M. Somatic mutagenesis and antimutagenesis in aging
rhabditis elegans. Proc. Natl. Acad. Sci. USA 90: 89058909, 1993. research. Mutat. Res. 350: 3541, 1996.
178a.LASS, A., S. AGARWAL, AND R. S. SOHAL. Mitochondrial ubiqui- 196. MARTIN, G. M., S. N. AUSTAD, AND T. E. JOHNSON. Genetic analy-
none homologues, superoxide radical generation, and longevity in sis of ageing: role of oxidative damage and environmental stresses.
different mammalian species. J. Biol. Chem. 272: 1919919204, Nature Genet. 13: 2534, 1996.
1997. 197. MARTINEZ, M., M. L. FERRANDIZ, A. DIEZ, AND J. MIQUEL. Deple-
179. LEE, C. M., S. S. CHUNG, J. M. KACZKOWSKI, R. WEINDRUCH, tion of cytosolic GSH decreases the ATP levels and viability of
AND J. M. AIKEN. Multiple mitochondrial DNA deletions associated synaptosomes from aged mice but not from young mice. Mech.
with age in skeletal muscle of rhesus monkeys. J. Gerontol. 48: Ageing Dev. 84: 7781, 1995.
201205, 1993. 198. MARTUS, H. J., M. E. DOLLE, J. A. GOSSEN, M. E. BOERRIGTER,
180. LEE, C. M., R. WEINDRUCH, AND J. M. AIKEN. Age-associated al- AND J. VIJG. Use of transgenic mouse models for studying somatic
terations of the mitochondrial genome. Free Radical Biol. Med. 22: mutations in aging. Mutat. Res. 338: 203213, 1995.
12591269, 1997. 199. MARZABADI, M. R., C. JONES, AND J. RYDSTROM. Indenoindole
181. LEVINE, R. L., D. GARLAND, C. N. OLIVER, A. AMICI, I. CLIMENT, depresses lipofuscin formation in cultured neonatal rat myocardial
A. G. LENZ, B. W. AHN, S. SHALTIEL, AND E. R. STADTMAN. Deter- cells. Mech. Ageing Dev. 80: 189197, 1995.
mination of carbonyl content in oxidatively modified proteins. 200. MARZABADI, M. R., R. S. SOHAL, AND U. T. BRUNK. Effect of ferric
Methods Enzymol. 186: 464478, 1990. iron and desferrioxamine on lipofuscin accumulation in cultured
182. LEZZA, A. M., D. BOFFOLI, S. SCACCO, P. CANTATORE, AND M. N. rat heart myocytes. Mech. Ageing Dev. 46: 145157, 1988.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


GADALETA. Correlation between mitochondrial DNA 4977-bp dele- 201. MARZABADI, M. R., R. S. SOHAL, AND U. T. BRUNK. Effect of
tion and respiratory chain enzyme activities in aging human skeletal alpha-tocopherol and some metal chelators on lipofuscin accumula-
muscles. Biochem. Biophys. Res. Commun. 205: 772779, 1994. tion in cultured neonatal rat cardiac myocytes. Ann. Cell Pathol.
183. LIEBETRAU, W., T. M. RUNGER, B. E. MEHLING, M. POOT, AND 2: 333346, 1990.
H. HOEHN. Mutagenic activity of ambient oxygen and mitomycin 202. MARZABADI, M. R., R. S. SOHAL, AND U. T. BRUNK. Mechanisms
C in Fanconis anaemia cells. Mutagenesis 12: 6977, 1997. of lipofuscinogenesis: effect of the inhibition of lysosomal protein-
183a.LIN, K., J. B. DORMAN, A. RODAN, AND C. KENYON. daf-16: an ases and lipases under varying concentrations of ambient oxygen
HNF-3/forkhead family member that can function to double the in cultured rat neonatal myocardial cells. APMIS 99: 416426, 1991.
life-span of Caenorhabditis elegans. Science 278: 13191322, 1997. 203. MATSUO, M. Oxygen dependency of life-span in the nematode.
184. LIOCHEV, S. I., AND I. FRIDOVICH. The role of O0 2 j in the produc- Comp. Biochem. Physiol. A Physiol. 105: 653658, 1993.
tion of jOH: in vitro and in vivo. Free Radical Biol. Med. 16: 29 203a.MATTSON, M. P. Advances fuel Alzheimers conundrum. Nature
33, 1994. Genet. 17: 254256, 1997.
185. LITHGOW, G. J., T. M. WHITE, S. MELOV, AND T. E. JOHNSON. 204. MCCORD, J. M., AND I. FRIDOVICH. Superoxide dismutase. An en-
Thermotolerance and extended life-span conferred by single-gene zymic function for erythrocuprein (hemocuprein). J. Biol. Chem.
mutations and induced by thermal stress. Proc. Natl. Acad. Sci. 244: 60496055, 1969.
USA 92: 75407544, 1995. 205. MEDVEDEV, Z. A. An attempt at a rational classification of theories
186. LIU, J., AND A. MORI. Age-associated changes in superoxide dismu- of ageing. Biol. Rev. Camb. Philos. Soc. 65: 375398, 1990.
tase activity, thiobarbituric acid reactivity and reduced glutathione 206. MELOV, S., G. J. LITHGOW, D. R. FISCHER, P. M. TEDESCO, AND
level in the brain and liver in senescence accelerated mice (SAM): T. E. JOHNSON. Increased frequency of deletions in the mitochon-
a comparison with ddY mice. Mech. Ageing Dev. 71: 2330, 1993. drial genome with age of Caenorhabditis elegans. Nucleic Acids
187. LIU, R. H., AND J. H. HOTCHKISS. Potential genotoxicity of chroni- Res. 23: 14191425, 1995.
cally elevated nitric oxide: a review. Mutat. Res. 339: 7389, 1995. 207. MELOV, S., J. M. SHOFFNER, A. KAUFMAN, AND D. C. WALLACE.
188. LOFT, S., A. ASTRUP, B. BUEMANN, AND H. E. POULSEN. Oxida- Marked increase in the number and variety of mitochondrial DNA
tive DNA damage correlates with oxygen consumption in humans. rearrangements in aging human skeletal muscle. Nucleic Acids Res.
FASEB J. 8: 534537, 1994. 23: 41224126, 1995.
189. LONNROT, K., T. METSA-KETELA, AND H. ALHO. The role of coen- 208. MENDIS, S., P. A. SOBOTKA, AND D. E. EULER. Pentane and iso-
zyme Q-10 in aging: a follow-up study on life-long oral supplementa- prene in expired air from humans: gas-chromatographic analysis
tion Q-10 in rats. Gerontology 41, Suppl. 2: 109120, 1995. of single breath. Clin. Chem. 40: 14851488, 1994.
190. LOPEZ-TORRES, M., R. PEREZ-CAMPO, C. ROJAS, S. CADENAS, 208a.MEYDANI, S. N., M. MEYDANI, J. B. BLUMBERG, L. S. LEKA, G.
AND G. BARJA. Maximum life span in vertebrates: relationship with SIBER, R. LOSZEWSKI, C. THOMPSON, M. C. PEDROSA, R. D.
liver antioxidant enzymes, glutathione system, ascorbate, urate, DIAMOND, AND B. D. STOLLAR. Vitamin E supplementation and
sensitivity to peroxidation, true malondialdehyde, in vivo H2O2 , and in vivo immune response in healthy elderly subjects. A randomized
basal and maximum aerobic capacity. Mech. Ageing Dev. 70: 177 controlled trial. JAMA 277: 13801386, 1997.
199, 1993. 209. MEYDANI, S. N., D. WU, M. S. SANTOS, AND M. G. HAYEK. Antioxi-
191. LOSCHEN, G., A. AZZI, C. RICHTER, AND L. FLOHE. Superoxide dants and immune response in aged persons: overview of present
radicals as precursors of mitochondrial hydrogen peroxide. FEBS evidence. Am. J. Clin. Nutr. 62, Suppl.: 1462S1476S, 1995.
Lett. 42: 6872, 1974. 210. MILLER, J. H. The relevance of bacterial mutators to understanding
192. LU, R., H. M. NASH, AND G. L. VERDINE. A mammalian DNA repair human cancer. Cancer Surv. 28: 141153, 1996.
enzyme that excises oxidatively damaged guanines maps to a locus 211. MINOTTI, G., AND S. D. AUST. The role of iron in the initiation of
frequently lost in lung cancer. Curr. Biol. 7: 397407, 1997. lipid peroxidation. Chem. Phys. Lipids. 44: 191208, 1987.
192a.LUCAS, D. T., AND L. I. SZWEDA. Cardiac reperfusion injury: aging, 212. MIQUEL, J. An update on the mitochondrial-DNA mutation hypoth-
lipid peroxidation, and mitochondrial dysfunction. Proc. Natl. esis of cell aging. Mutat. Res. 275: 209216, 1992.
Acad. Sci. USA 95: 510514, 1998. 213. MIQUEL, J., A. C. ECONOMOS, J. FLEMING, AND J. J. JOHNSON.
193. LUFT, R. The development of mitochondrial medicine. Proc. Natl. Mitochondrial role in cell aging. Exp. Gerontol. 15: 575591, 1980.
Acad. Sci. USA 91: 87318738, 1994. 214. MIQUEL, J., M. L. FERRANDIZ, E. DE JUAN, I. SEVILA, AND M.
193a.MARBOIS, B. N., AND C. F. CLARKE. The COQ7 gene encodes MARTINEZ. N-acetylcysteine protects against age-related decline
a protein in Saccharomyces cerevisiae necessary for ubiquinone of oxidative phosphorylation in liver mitochondria. Eur. J. Phar-
biosynthesis. J. Biol. Chem. 271: 29953004, 1996. macol. 292: 333335, 1995.
193b.MARKESBERY, W. R. Oxidative stress hypothesis in Alzheimers 215. MO, J. Q., D. G. HOM, AND J. K. ANDERSEN. Decreases in protec-
disease. Free Radical Biol. Med. 23: 134147, 1997. tive enzymes correlates with increased oxidative damage in the
194. MARNETT, L. J., H. K. HURD, M. C. HOLLSTEIN, D. E. LEVIN, H. aging mouse brain. Mech. Ageing Dev. 81: 7382, 1995.
ESTERBAUER, AND B. N. AMES. Naturally occurring carbonyl 216. MONNIER, V. M., AND A. CERAMI. Nonenzymatic browning in vivo:

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


578 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

possible process for aging of long-lived proteins. Science 211: 491 systems in oxidative stress: the free-radical theory of aging revis-
493, 1981. ited. Gerontology 37: 166180, 1991.
217. MONNIER, V. M., M. GLOMB, A. ELGAWISH, AND D. R. SELL. The 240. PACIFICI, R. E., Y. KONO, AND K. J. DAVIES. Hydrophobicity as
mechanism of collagen cross-linking in diabetes: a puzzle nearing the signal for selective degradation of hydroxyl radical-modified
resolution. Diabetes 45, Suppl. 3: S67S72, 1996. hemoglobin by the multicatalytic proteinase complex, proteasome.
218. MORLEY, A. A. The somatic mutation theory of ageing. Mutat. Res. J. Biol. Chem. 268: 1540515411, 1993.
338: 1923, 1995. 241. PACKER, L., AND K. FUEHR. Low oxygen concentration extends
219. MORRIS, J. Z., H. A. TISSENBAUM, AND G. RUVKUN. A phosphati- the life span of cultured human diploid cells. Nature 267: 423425,
dylinositol-3-OH kinase family member regulating longevity and 1977.
diapause in Caenorhabditis elegans. Nature 382: 536539, 1996. 242. PACKER, L., AND J. R. SMITH. Extension of the life span of cultured
220. MOSLEN, M. T. Reactive oxygen species in normal physiology, cell normal human diploid cells by vitamin E: a reevaluation. Proc. Natl.
injury and phagocytosis. Adv. Exp. Med. Biol. 366: 1727, 1994. Acad. Sci. USA 74: 16401641, 1977.
221. MULLAART, E., P. H. LOHMAN, F. BERENDS, AND J. VIJG. DNA 243. PALLOTTI, F., X. CHEN, E. BONILLA, AND E. A. SCHON. Evidence
damage metabolism and aging. Mutat. Res. 237: 189210, 1990. that specific mtDNA point mutations may not accumulate in skele-
222. MULLER, H. J. Mitochondria and ageing. Brain Pathol. 2: 149158, tal muscle during normal human aging. Am. J. Hum. Genet. 59:
1992. 591602, 1996.
223. NAGLEY, P., I. R. MACKAY, A. BAUMER, R. J. MAXWELL, F. VAIL- 244. PARK, J. W., C. H. CHOI, M. S. KIM, AND M. H. CHUNG. Oxidative
LANT, Z. X. WANG, C. ZHANG, AND A. W. LINNANE. Mitochondrial status in senescence-accelerated mice. J. Gerontol. 51: 337345,
DNA mutation associated with aging and degenerative disease. 1996.
Ann. NY Acad. Sci. 673: 92102, 1992. 245. PARSONNET, J., G. D. FRIEDMAN, D. P. VANDERSTEEN, Y.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


224. NAKANO, M., AND S. GOTOH. Accumulation of cardiac lipofuscin CHANG, J. H. VOGELMAN, N. ORENTREICH, AND R. K. SIBLEY.
depends on metabolic rate of mammals. J. Gerontol. 47: 126129, Helicobacter pylori infection and the risk of gastric carcinoma. N.
1992. Engl. J. Med. 325: 11271131, 1991.
225. NAKANO, M., T. MIZUNO, AND S. GOTOH. Accumulation of cardiac 246. PEARL, R. The Rate of Living. London: Univ. of London Press,
lipofuscin in crab-eating monkeys (Macaca fasicularis): the same 1928.
rate of lipofuscin accumulation in several species of primates. 247. PEREZ-CAMPO, R., M. LOPEZ-TORRES, C. ROJAS, S. CADENAS,
Mech. Ageing Dev. 66: 243248, 1993. AND G. BARJA. Longevity and antioxidant enzymes, non-enzymatic
226. NAKANO, M., T. MIZUNO, H. KATOH, AND S. GOTOH. Age-related antioxidants and oxidative stress in the vertebrate lung: a compara-
accumulation of lipofuscin in myocardium of Japanese monkey tive study. J. Comp. Physiol. B Biochem. Syst. Environ. Physiol.
(Macaca fuscata). Mech. Ageing Dev. 49: 4148, 1989. 163: 682689, 1994.
227. NOHL, H. Involvement of free radicals in ageing: a consequence or 248. PEREZ-CAMPO, R., M. LOPEZ-TORRES, C. ROJAS, S. CADENAS,
cause of senescence. Br. Med. Bull. 49: 653667, 1993. G. BARJA, AND D. QUIROGA. Lung glutathione reductase induction
228. NOHL, H., L. GILLE, K. SCHONHEIT, AND Y. LIU. Conditions in aging catalase-depleted frogs correlates with early survival
allowing redox-cycling ubisemiquinone in mitochondria to estab- throughout the life span. Mech. Ageing Dev. 67: 115127, 1993.
lish a direct redox couple with molecular oxygen. Free Radical 248a.PFLAUM, M., O. WILL, AND B. EPE. Determination of steady-state
Biol. Med. 20: 207213, 1996. levels of oxidative DNA base modifications in mammalian cells
229. NOHL, H., AND D. HEGNER. Do mitochondria produce oxygen radi- by means of repair endonucleases. Carcinogenesis 18: 22252231,
cals in vivo? Eur. J. Biochem. 82: 563567, 1978. 1997.
230. OCKNER, R. K., R. M. KAIKAUS, AND N. M. BASS. Fatty-acid metab- 249. PODEROSO, J. J., M. C. CARRERAS, C. LISDERO, N. RIOBO, F.
olism and the pathogenesis of hepatocellular carcinoma: review SCHOPFER, AND A. BOVERIS. Nitric oxide inhibits electron trans-
and hypothesis. Hepatology 18: 669676, 1993. fer and increases superoxide radical production in rat heart mito-
231. OHSHIMA, H., AND H. BARTSCH. Chronic infections and inflamma- chondria and submitochondrial particles. Arch. Biochem. Biophys.
tory processes as cancer risk factors: possible role of nitric oxide 328: 8592, 1996.
in carcinogenesis. Mutat. Res. 305: 253264, 1994. 250. POOT, M. Oxidants and antioxidants in proliferative senescence.
232. OIKAWA, I., AND P. M. NOVIKOFF. Catalase-negative peroxisomes:
Mutat. Res. 256: 177189, 1991.
transient appearance in rat hepatocytes during liver regeneration
251. PORTA, E. A. Advances in age pigment research. Arch. Gerontol.
after partial hepatectomy. Am. J. Pathol. 146: 673687, 1995.
Geriatr. 12: 303320, 1991.
233. OLIVER, C. N., B. W. AHN, E. J. MOERMAN, S. GOLDSTEIN, AND
252. POTTER, J. D. Beta-carotene and the role of intervention studies.
E. R. STADTMAN. Age-related changes in oxidized proteins. J. Biol.
Chem. 262: 54885491, 1987. Cancer Lett. 114: 329331, 1997.
234. OLIVER, C. N., P. E. STARKE-REED, E. R. STADTMAN, G. J. LIU, 252a.PROFT, M., P. KOTTER, D. HEDGES, N. BOJUNGA, AND K. D.
J. M. CARNEY, AND R. A. FLOYD. Oxidative damage to brain pro- ENTIAN. CAT5, a new gene necessary for derepression of gluco-
teins, loss of glutamine synthetase activity, and production of free neogenic enzymes in Saccharomyces cerevisiae. EMBO J. 14:
radicals during ischemia/reperfusion-induced injury to gerbil brain. 61166126, 1995.
Proc. Natl. Acad. Sci. USA 87: 51445147, 1990. 253. RAO, G., E. XIA, M. J. NADAKAVUKAREN, AND A. RICHARDSON.
235. OLLINGER, K., AND U. T. BRUNK. Cellular injury induced by oxida- Effect of dietary restriction on the age-dependent changes in the
tive stress is mediated through lysosomal damage. Free Radical expression of antioxidant enzymes in rat liver. J. Nutr. 120: 602
Biol. Med. 19: 565574, 1995. 609, 1990.
236. OMENN, G. S., G. E. GOODMAN, M. D. THORNQUIST, J. BALMES, 254. RAO, G., E. XIA, AND A. RICHARDSON. Effect of age on the expres-
M. R. CULLEN, A. GLASS, J. P. KEOGH, F. L. MEYSKENS, JR., B. sion of antioxidant enzymes in male Fischer F344 rats. Mech. Age-
VALANIS, J. H. WILLIAMS, JR., S. BARNHART, M. G. CHERNIACK, ing Dev. 53: 4960, 1990.
C. A. BRODKIN, AND S. HAMMAR. Risk factors for lung cancer and 255. REAUME, A. G., J. L. ELLIOTT, E. K. HOFFMAN, N. W. KOWALL,
for intervention effects in CARET, the beta-carotene and retinol R. J. FERRANTE, D. F. SIWEK, H. M. WILCOX, D. G. FLOOD, M. F.
efficacy trial. J. Natl. Cancer Inst. 88: 15501559, 1996. BEAL, R. H. BROWN, JR., R. W. SCOTT, AND W. D. SNIDER. Motor
237. ORR, W. C., AND R. S. SOHAL. Extension of life-span by overex- neurons in Cu/Zn superoxide dismutase-deficient mice develop nor-
pression of superoxide dismutase and catalase in Drosophila mela- mally but exhibit enhanced cell death after axonal injury. Nature
nogaster. Science 263: 11281130, 1994. Genet. 13: 4347, 1996.
238. OZAWA, T. Mechanism of somatic mitochondrial DNA mutations 255a.REED, J. C. Cytochomre c: cant live with itcant live without it.
associated with age and diseases. Biochim. Biophys. Acta 1271: Cell 91: 559562, 1997.
177189, 1995. 256. REID, T. M., D. I. FEIG, AND L. A. LOEB. Mutagenesis by metal-
238a.OZAWA, T. Genetic and functional changes in mitochondria associ- induced oxygen radicals. Environ Health Perspect. 102, Suppl. 3:
ated with aging. Physiol. Rev. 77: 425464, 1997. 5761, 1994.
239. PACIFICI, R. E., AND K. J. DAVIES. Protein, lipid and DNA repair 257. REITER, R. J., M. I. PABLOS, T. T. AGAPITO, AND J. M. GUER-

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 579

RERO. Melatonin in the context of the free radical theory of aging. 279. SCHMIDT, A. M., O. HORI, J. X. CHEN, J. F. LI, J. CRANDALL, J.
Ann. NY Acad. Sci. 786: 362378, 1996. ZHANG, R. CAO, S. D. YAN, J. BRETT, AND D. STERN. Advanced
258. RICHTER, C. Reactive oxygen and DNA damage in mitochondria. glycation end products interacting with their endothelial receptor
Mutat. Res. 275: 249255, 1992. induce expression of vascular cell adhesion molecule-1 (VCAM-
259. RICHTER, C. Oxidative damage to mitochondrial DNA and its rela- 1) in cultured human endothelial cells and in mice. A potential
tionship to ageing. Int. J. Biochem. Cell Biol. 27: 647653, 1995. mechanism for the accelerated vasculopathy of diabetes. J. Clin.
260. RICHTER, C., J. W. PARK, AND B. N. AMES. Normal oxidative dam- Invest. 96: 13951403, 1995.
age to mitochondrial and nuclear DNA is extensive. Proc. Natl. 280. SCHMIDT, A. M., S. D. YAN, J. BRETT, R. MORA, R. NOWYGROD,
Acad. Sci. USA 85: 64656467, 1988. AND D. STERN. Regulation of human mononuclear phagocyte mi-
261. RIVETT, A. J. Intracellular protein degradation. Essays Biochem. gration by cell surface-binding proteins for advanced glycation end
25: 3981, 1990. products. J. Clin. Invest. 91: 21552168, 1993.
262. ROBINSON, J. M., AND J. A. BADWEY. Production of active oxygen 281. SCHULZ, J. B., R. T. MATTHEWS, D. R. HENSHAW, AND M. F.
species by phagocytic leukocytes. Immunol. Ser. 60: 159178, 1994. BEAL. Neuroprotective strategies for treatment of lesions pro-
263. RODRIGUEZ-MARTINEZ, M. A., AND A. RUIZ-TORRES. Homeosta- duced by mitochondrial toxins: implications for neurodegenerative
sis between lipid peroxidation and antioxidant enzyme activities diseases. Neuroscience 71: 10431048, 1996.
in healthy human aging. Mech. Ageing Dev. 66: 213222, 1992. 282. SCHWARZE, S. R., C. M. LEE, S. S. CHUNG, E. B. ROECKER, R.
264. ROSE, M. R. Laboratory evolution of postponed senescence in Dro- WEINDRUCH, AND J. M. AIKEN. High levels of mitochondrial DNA
sophila melanogaster. Evolution 38: 10041010, 1984. deletions in skeletal muscle of old rhesus monkeys. Mech. Ageing
265. ROSE, M. R., AND C. E. FINCH. Genetics and evolution of aging. Dev. 83: 91101, 1995.
In: Contemporary Issues in Genetics and Evolution. Dordrecht, 283. SELKOE, D. J. Cell biology of the amyloid beta-protein precursor
and the mechanism of Alzheimers disease. Annu. Rev. Cell Biol.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


Germany: Kluwer, 1994, vol 3.
266. ROSENQUIST, T. A., D. O. ZHARKOV, AND A. P. GROLLMAN. Clon- 10: 373403, 1994.
ing and characterization of a mammalian 8-oxoguanine DNA glyco- 284. SELL, D. R., M. A. LANE, W. A. JOHNSON, E. J. MASORO, O. B.
sylase. Proc. Natl. Acad. Sci. USA 94: 74297434, 1997. MOCK, K. M. REISER, J. F. FOGARTY, R. G. CUTLER, D. K. IN-
267. ROSSMAN, T. G., E. I. GONCHAROVA, A. NADAS, AND N. DOLZ- GRAM, G. S. ROTH, AND V. M. MONNIER. Longevity and the genetic
HANSKAYA. Chinese hamster cells expressing antisense to metallo- determination of collagen glycoxidation kinetics in mammalian se-
thionein become spontaneous mutators. Mutat. Res. 373: 7585, nescence. Proc. Natl. Acad. Sci. USA 93: 485490, 1996.
1997. 285. SEN, C. K., AND L. PACKER. Antioxidant and redox regulation of
268. RUBNER, M. Das problem der Lebensdauer und seine Bezie- gene transcription. FASEB J. 10: 709720, 1996.
hungen zu Wachstum und Ernehrung. Munich: Oldenburg, 1908. 286. SERRANO, M., A. W. LIN, M. E. MCCURRACH, D. BEACH, AND
269. SAGER, R. Senescence as a mode of tumor suppression. Environ. S. W. LOWE. Oncogenic ras provokes premature cell senescence
Health Perspect. 93: 5962, 1991. associated with accumulation of p53 and p16INK4a. Cell 88: 593
270. SAHAKIAN, J. A., L. I. SZWEDA, B. FRIGUET, K. KITANI, AND R. L. 602, 1997.
LEVINE. Aging of the liver: proteolysis of oxidatively modified glu- 287. SHIGENAGA, M. K., C. J. GIMENO, AND B. N. AMES. Urinary 8-
tamine synthetase. Arch. Biochem. Biophys. 318: 411417, 1995. hydroxy-2*-deoxyguanosine as a biological marker of in vivo oxida-
270a.SALAZAR, J. M., AND B. VAN HOUTEN. Preferential mitochondrial tive DNA damage. Proc. Natl. Acad. Sci. USA 86: 96979701, 1989.
DNA injury caused by glucose oxidase as a steady generator of 288. SHIGENAGA, M. K., T. M. HAGEN, AND B. N. AMES. Oxidative
hydrogen peroxide in human fibroblasts. Mutat. Res. 385: 139149, damage and mitochondrial decay in aging. Proc. Natl. Acad. Sci.
1997. USA 91: 1077110778, 1994.
271. SALGANIK, R. I., I. G. SHABALINA, N. A. SOLOVYOVA, N. G. KO- 289. SHMOOKLER REIS, R. J., AND R. H. N. EBERT. Genetics of aging:
LOSOVA, V. N. SOLOVYOV, AND A. R. KOLPAKOV. Impairment of current animal models. Exp. Gerontol. 31: 6981, 1996.
respiratory functions in mitochondria of rats with an inherited hy- 290. SHOFFNER, J. M., M. D. BROWN, A. TORRONI, M. T. LOTT, M. F.
perproduction of free radicals. Biochem. Biophys. Res. Commun. CABELL, S. S. MIRRA, M. F. BEAL, C. C. YANG, M. GEARING, AND
205: 180185, 1994. R. SALVO. Mitochondrial DNA variants observed in Alzheimer dis-
272. SALGANIK, R. I., N. A. SOLOVYOVA, S. I. DIKALOV, O. N. GRIS- ease and Parkinson disease patients. Genomics 17: 171184, 1993.
HAEVA, L. A. SEMENOVA, AND A. V. POPOVSKY. Inherited en- 290a.SLATER, A. F., C. S. NOBEL, AND S. ORRENIUS. The role of intra-
hancement of hydroxyl radical generation and lipid peroxidation cellular oxidants in apoptosis. Biochim. Biophys. Acta 1271: 59
in the S strain rats results in DNA rearrangements, degenerative 62, 1995.
diseases, and premature aging. Biochem. Biophys. Res. Commun. 291. SLUPSKA, M. M., C. BAIKALOV, W. M. LUTHER, J. H. CHIANG,
199: 726733, 1994. Y. F. WEI, AND J. H. MILLER. Cloning and sequencing a human
272a.SASTRE, J., A. MILLAN, J. GARCIA DE LA ASUNCION, R. PLA, G. homolog (hMYH) of the Escherichia coli mutY gene whose function
JUAN, F. V. PALLARDO, E. OCONNOR, J. A. MARTIN, M.-T. DROY- is required for the repair of oxidative DNA damage. J. Bacteriol.
LEFAIX, AND J. VINA. A Ginko Biloba extract (EGb761) prevents 178: 38853892, 1996.
mitochondrial aging by protecting against oxidative stress. Free 291a.SMEAL, T., AND L. GUARENTE. Mechanisms of cellular senes-
Radical Biol. Med. 24: 298304, 1998. cence. Curr. Opin. Genet. Dev. 7: 281287, 1997.
273. SASTRE, J., F. V. PALLARDO, AND J. VINA. Glutathione, oxidative 292. SMITH, D. W. E. Centenarians: human longevity outliers. Gerontol-
stress, and aging. Age 19: 129139, 1996. ogist 37: 200207, 1997.
274. SAWADA, M., AND J. C. CARLSON. Biochemical changes associated 292a.SMITH, M. A., P. L. HARRIS, L. M. SAYRE, AND G. PERRY. Iron
with the mechanism controlling superoxide radical formation in accumulation in Alzheimer disease is a source of redox-generated
the aging rotifer. J. Cell. Biochem. 44: 153165, 1990. free radicals. Proc. Natl. Acad. Sci. USA 94: 98669868, 1997.
275. SCHAPIRA, A. H. Oxidative stress in Parkinsons disease. Neuropa- 292b.SMITH, M. A., G. PERRY, P. L. RICHEY, L. M. SAYRE, V. E. ANDER-
thol. Appl. Neurobiol. 21: 39, 1995. SON, M. F. BEAL, AND N. KOWALL. Oxidative damage in Alzhei-
276. SCHAPIRA, A. H., AND J. M. COOPER. Mitochondrial function in mers. Nature 382: 120121, 1996.
neurodegeneration and ageing. Mutat. Res. 275: 133143, 1992. 293. SMITH, M. A., L. M. SAYRE, V. M. MONNIER, AND G. PERRY. Radi-
277. SCHMIDT, A. M., O. HORI, J. BRETT, S. D. YAN, J. L. WAUTIER, cal ageing in Alzheimers disease. Trends Neurosci. 18: 172176,
AND D. STERN. Cellular receptors for advanced glycation end prod- 1995.
ucts. Implications for induction of oxidant stress and cellular dys- 294. SOCCI, D. J., B. M. CRANDALL, AND G. W. ARENDASH. Chronic
function in the pathogenesis of vascular lesions. Arterioscler. antioxidant treatment improves the cognitive performance of aged
Thromb. 14: 15211528, 1994. rats. Brain Res. 693: 8894, 1995.
278. SCHMIDT, A. M., O. HORI, R. CAO, S. D. YAN, J. BRETT, J. L. 295. SOHAL, R. S. Metabolic rate and life span. In: Cellular Aging: Con-
WAUTIER, S. OGAWA, K. KUWABARA, M. MATSUMOTO, AND D. cepts and Mechanisms, edited by R. Witler. Basel: Karger, 1976, p.
STERN. RAGE: a novel cellular receptor for advanced glycation 2540.
end products. Diabetes 45, Suppl. 3: S77S80, 1996. 296. SOHAL, R. S. Age Pigments. Amsterdam: Elsevier, 1981.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


580 KENNETH B. BECKMAN AND BRUCE N. AMES Volume 78

297. SOHAL, R. S. The free radical hypothesis of aging: an appraisal of rial defenses against oxidative stress. Trends Genet. 6: 363368,
the current status. Aging Clin. Exp. Res. 5: 317, 1993. 1990.
298. SOHAL, R. S., A. AGARWAL, S. AGARWAL, AND W. C. ORR. Simul- 321. SUZUKI, Y. J., H. J. FORMAN, AND A. SEVANIAN. Oxidants as stim-
taneous overexpression of copper- and zinc-containing superoxide ulators of signal transduction. Free Radical Biol. Med. 22: 269
dismutase and catalase retards age-related oxidative damage and 285, 1997.
increases metabolic potential in Drosophila melanogaster. J. Biol. 322. SZWEDA, L. I., K. UCHIDA, L. TSAI, AND E. R. STADTMAN. Inacti-
Chem. 270: 1567115674, 1995. vation of glucose-6-phosphate dehydrogenase by 4-hydroxy-2-non-
299. SOHAL, R. S., S. AGARWAL, A. DUBEY, AND W. C. ORR. Protein enal. Selective modification of an active-site lysine. J. Biol. Chem.
oxidative damage is associated with life expectancy of houseflies. 268: 33423347, 1993.
Proc. Natl. Acad. Sci. USA 90: 72557259, 1993. 322a.TANAKA, M., J.-S. GONG, J. ZHANG, M. YONEDA, AND K. YAGI.
300. SOHAL, R. S., S. AGARWAL, AND B. H. SOHAL. Oxidative stress Mitochondrial genotype associated with longevity. Lancet 351: 185
and aging in the Mongolian gerbil (Meriones unguiculatus). Mech. 186, 1998.
Ageing Dev. 81: 1525, 1995. 323. TANHAUSER, S. M., AND P. J. LAIPIS. Multiple deletions are detect-
301. SOHAL, R. S., L. ARNOLD, AND W. C. ORR. Effect of age on super- able in mitochondrial DNA of aging mice. J. Biol. Chem. 270:
oxide dismutase, catalase, glutathione reductase, inorganic perox- 2476924775, 1995.
ides, TBA-reactive material, GSH/GSSG, NADPH/NADP/ and 324. TAPPEL, A. L. Lipid peroxidation and fluorescent molecular dam-
NADH/NAD/ in Drosophila melanogaster. Mech. Ageing Dev. 56: age to membranes. In: Pathobiology of Cell Membranes, edited by
223235, 1990. B. F. Trump and A. U. Arstila. New York: Academic, 1975, p. 145
302. SOHAL, R. S., L. A. ARNOLD, AND B. H. SOHAL. Age-related 170.
changes in antioxidant enzymes and prooxidant generation in tis- 325. TCHOU, J., H. KASAI, S. SHIBUTANI, M. H. CHUNG, J. LAVAL,
sues of the rat with special reference to parameters in two insect A. P. GROLLMAN, AND S. NISHIMURA. 8-Oxoguanine (8-hydroxy-

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


species. Free Radical Biol. Med. 9: 495500, 1990. guanine) DNA glycosylase and its substrate specificity. Proc. Natl.
303. SOHAL, R. S., AND A. DUBEY. Mitochondrial oxidative damage, Acad. Sci. USA 88: 46904694, 1991.
hydrogen peroxide release, and aging. Free Radical Biol. Med. 16: 326. THORPE, S. R., AND J. W. BAYNES. Role of the Maillard reaction
621626, 1994. in diabetes mellitus and diseases of aging. Drugs Aging 9: 6977,
304. SOHAL, R. S., H. H. KU, AND S. AGARWAL. Biochemical correlates 1996.
of longevity in two closely related rodent species. Biochem. Bio- 327. TSUCHIDA, M., T. MIURA, AND K. AIBARA. Lipofuscin and lipofus-
phys. Res. Commun. 196: 711, 1993. cin-like substances. Chem. Phys. Lipids 44: 297325, 1987.
305. SOHAL, R. S., H. H. KU, S. AGARWAL, M. J. FORSTER, AND H. LAL. 328. TYLER, R. H., H. BRAR, M. SINGH, A. LATORRE, J. L. GRAVES,
Oxidative damage, mitochondrial oxidant generation and antioxi- L. D. MUELLER, M. R. ROSE, AND F. J. AYALA. The effect of super-
dant defenses during aging and in response to food restriction in oxide dismutase alleles on aging in Drosophila. Genetica 91: 143
the mouse. Mech. Ageing Dev. 74: 121133, 1994. 149, 1993.
306. SOHAL, R. S., M. R. MARZABADI, D. GALARIS, AND U. T. BRUNK. 329. UEJIMA, Y., Y. FUKUCHI, S. TERAMOTO, R. TABATA, AND H.
Effect of ambient oxygen concentration on lipofuscin accumulation ORIMO. Age changes in visceral content of glutathione in the senes-
in cultured rat heart myocytesa novel in vitro model of lipofusci- cence accelerated mouse (SAM). Mech. Ageing Dev. 67: 129139,
nogenesis. Free Radical Biol. Med. 6: 2330, 1989. 1993.
307. SOHAL, R. S., AND W. C. ORR. Relationship between antioxidants, 330. VANFLETEREN, J. R. Oxidative stress and ageing in Caenorhab-
prooxidants, and the aging process. Ann. NY Acad. Sci. 663: 74 ditis elegans. Biochem. J. 292: 605608, 1993.
84, 1992. 330a.VAZIRI, H., M. D. WEST, R. C. ALLSOPP, T. S. DAVISON, Y. S. WU,
308. SOHAL, R. S., AND B. H. SOHAL. Hydrogen peroxide release by C. H. ARROWSMITH, G. G. POIRIER, AND S. BENCHIMOL. ATM-
mitochondria increases during aging. Mech. Ageing Dev. 57: 187 dependent telomere loss in aging human diploid fibroblasts and
202, 1991. DNA damage lead to the post-translational activation of p53 protein
309. SOHAL, R. S., B. H. SOHAL, AND U. T. BRUNK. Relationship be- involving poly(ADP-ribose) polymerase. EMBO J. 16: 60186033,
tween antioxidant defenses and longevity in different mammalian 1997.
species. Mech. Ageing Dev. 53: 217227, 1990. 331. VERCELLOTTI, G. M. A balanced budgetevaluating the iron
310. SOHAL, R. S., B. H. SOHAL, AND W. C. ORR. Mitochondrial superox- economy. Clin. Chem. 42: 657, 1996.
ide and hydrogen peroxide generation, protein oxidative damage, 332. VIJG, J., AND J. A. GOSSEN. Somatic mutations and cellular aging.
and longevity in different species of flies. Free Radical Biol. Med. Comp. Biochem. Physiol B Biochem. Mol. Physiol. 104: 429437,
19: 499504, 1995. 1993.
311. SOHAL, R. S., I. SVENSSON, AND U. T. BRUNK. Hydrogen peroxide 333. VON ZGLINICKI, T., E. NILSSON, W. D. DOCKE, AND U. T. BRUNK.
production by liver mitochondria in different species. Mech. Ageing Lipofuscin accumulation and ageing of fibroblasts. Gerontology 41,
Dev. 53: 209215, 1990. Suppl. 2: 95108, 1995.
312. SOHAL, R. S., AND R. WEINDRUCH. Oxidative stress, caloric re- 334. VON ZGLINICKI, T., G. SARETZKI, W. DOCKE, AND C. LOTZE.
striction, and aging. Science 273: 5963, 1996. Mild hyperoxia shortens telomeres and inhibits proliferation of fi-
313. SOONG, N. W., D. R. HINTON, G. CORTOPASSI, AND N. ARNHEIM. broblasts: a model for senescence? Exp. Cell Res. 220: 186193,
Mosaicism for a specific somatic mitochondrial DNA mutation in 1995.
adult human brain. Nature Genet. 2: 318323, 1992. 335. WACHSMAN, J. T. The beneficial effects of dietary restriction: re-
314. SPRINGFIELD, J. R., AND M. D. LEVITT. Pitfalls in the use of breath duced oxidative damage and enhanced apoptosis. Mutat. Res. 350:
pentane measurements to assess lipid peroxidation. J. Lipid Res. 2534, 1996.
35: 14971504, 1994. 336. WALLACE, D. C., J. M. SHOFFNER, I. TROUNCE, M. D. BROWN,
315. STADTMAN, E. R. Protein oxidation and aging. Science 257: 1220 S. W. BALLINGER, D. M. CORRAL, T. HORTON, A. S. JUN, AND
1224, 1992. M. T. LOTT. Mitochondrial DNA mutations in human degenerative
316. STADTMAN, E. R. Role of oxidized amino acids in protein break- diseases and aging. Biochim. Biophys. Acta 1271: 141151, 1995.
down and stability. Methods Enzymol. 258: 379393, 1995. 336a.WALLACE, D. C., C. STUGARD, D. MURDOCH, T. SCHURR, AND
317. STADTMAN, E. R., AND C. N. OLIVER. Metal-catalyzed oxidation M. D. BROWN. Ancient mtDNA sequences in the human nuclear
of proteins. Physiological consequences. J. Biol. Chem. 266: 2005 genome: a potential source of errors in identifying pathogenic muta-
2008, 1991. tions. Proc. Natl. Acad. Sci. USA 94: 1490014905, 1997.
318. STARKE-REED, P. E., AND C. N. OLIVER. Protein oxidation and 337. WANG, Y. J., Y. S. HO, M. J. LO, AND J. K. LIN. Oxidative modifica-
proteolysis during aging and oxidative stress. Arch. Biochem. Bio- tion of DNA bases in rat liver and lung during chemical carcinogene-
phys. 275: 559567, 1989. sis and aging. Chem. Biol. Interact. 94: 135145, 1995.
319. STEIN, G., AND J. WEISS. Chemical effects of ionizing radiations. 338. WARNER, H. R. Superoxide dismutase, aging, and degenerative dis-
Nature 161: 650, 1948. ease. Free Radical Biol. Med. 17: 249258, 1994.
320. STORZ, G., L. A. TARTAGLIA, S. B. FARR, AND B. N. AMES. Bacte- 338a.WARNER, H. R., R. J. HODES, AND K. POCINKI. What does cell

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev


April 1998 FREE RADICAL THEORY OF AGING 581

death have to do with aging? J. Am. Geriatr. Soc. 45: 11401146, ment of free radical formation due to a decrease in Km for hydrogen
1997. peroxide. Proc. Natl. Acad. Sci. USA 93: 57095714, 1996.
338b.WARNER, H. R., AND T. E. JOHNSON. Parsing age, mutations and 347. YIN, D. Age Pigments and the Biochemical Basis of Their Forma-
time. Nature Genet. 17: 368370, 1997. tion. Linkuping, Sweden: Linkuping University, 1995.
339. WEI, Y. H. Mitochondrial DNA alterations as aging-associated mo- 348. YIN, D. Biochemical basis of lipofuscin, ceroid, and age pigment-
lecular events. Mutat. Res. 275: 145155, 1992. like fluorophores. Free Radical Biol. Med. 21: 871888, 1996.
339a.WEINDRUCH, R., AND R. S. SOHAL. Seminars in medicine of the 349. YONEDA, T., M. HIRAMATSU, M. SAKAMOTO, K. TOGASAKI, M.
Beth Israel Deaconess Medical Center. Caloric intake and aging. KOMATSU, AND K. YAMAGUCHI. Antioxidant effects of beta cate-
N. Engl. J. Med. 337: 986994, 1997. chin. Biochem. Mol. Biol. Int. 35: 9951008, 1995.
340. WEINDRUCH, R., R. L. WALFORD, S. FLIGIEL, AND D. GUTHRIE. 350. YOUNGMAN, L. D., J. Y. PARK, AND B. N. AMES. Protein oxidation
The retardation of aging in mice by dietary restriction: longevity, associated with aging is reduced by dietary restriction of protein
cancer, immunity and lifetime energy intake. J. Nutr. 116: 641 or calories (published erratum appears in Proc. Natl. Acad. Sci.
654, 1986. USA 89: 11107, 1992). Proc. Natl. Acad. Sci. USA 89: 91129116,
341. WILLIAMS, G. C., AND R. M. NESSE. The dawn of Darwinian medi- 1992.
cine. Q. Rev. Biol. 66: 122, 1991. 351. YU, B. P. Cellular defenses against damage from reactive oxygen
342. WOOD, P. M. The potential diagram for oxygen at pH 7. Biochem. species. Physiol. Rev. 74: 139162, 1994.
J. 253: 287289, 1988. 352. YU, B. P. Aging and oxidative stress: modulation by dietary restric-
343. YAGI, K., S. KOMURA, Y. SASAGURI, K. YOSHINO, AND N. OHISHI. tion. Free Radical Biol. Med. 5: 651668, 1996.
Atherogenic change in the thoracic aorta of the senescence-acceler- 353. YU, B. P., AND R. YANG. Critical evaluation of the free radical theory
ated mouse. Atherosclerosis 118: 233236, 1995. of aging. A proposal for the oxidative stress hypothesis. Ann. NY
343a.YAKES, F. M., AND B. VAN HOUTEN. Mitochondrial DNA damage Acad. Sci. 786: 111, 1996.

Downloaded from http://physrev.physiology.org/ by 10.220.33.4 on May 21, 2017


is more extensive and persists longer than nuclear DNA damage 354. YU, C. E., J. OSHIMA, Y. H. FU, E. M. WIJSMAN, F. HISAMA,
in human cells following oxidative stress. Proc. Natl. Acad. Sci. R. ALISCH, S. MATTHEWS, J. NAKURA, T. MIKI, S. OUAIS,
USA 94: 514519, 1997. G. M. MARTIN, J. MULLIGAN, AND G. D. SCHELLENBERG. Po-
343b.YAN, L. J., R. L. LEVINE, AND R. S. SOHAL. Oxidative damage sitional cloning of the Werners syndrome gene. Science 272:
during aging targets mitochondrial aconitase. Proc. Natl. Acad. Sci. 258 262, 1996.
USA 94: 1116811172, 1997. 355. YUAN, H., T. KANEKO, AND M. MATSUO. Relevance of oxidative
344. YAN, S. D., X. CHEN, J. FU, M. CHEN, H. ZHU, A. ROHER, T. stress to the limited replicative capacity of cultured human diploid
SLATTERY, L. ZHAO, M. NAGASHIMA, J. MORSER, A. MIGHELI, cells: the limit of cumulative population doublings increases under
P. NAWROTH, D. STERN, AND A. M. SCHMIDT. RAGE and amyloid- low concentrations of oxygen and decreases in response to amino-
beta peptide neurotoxicity in Alzheimers disease. Nature 382: 685 triazole. Mech. Ageing Dev. 81: 159168, 1995.
691, 1996. 355a.ZAMZAMI, N., T. HIRSCH, B. DALLAPORTA, P. X. PETIT, AND G.
345. YELINOVA, V., Y. GLAZACHEV, V. KHRAMTSOV, L. KUDRYAS- KROEMER. Mitochondrial implication in accidental and pro-
HOVA, V. RYKOVA, AND R. SALGANIK. Studies of human and rat grammed cell death: apoptosis and necrosis. J. Bioenerg. Bio-
blood under oxidative stress: changes in plasma thiol level, antioxi- membr. 29: 185193, 1997.
dant enzyme activity, protein carbonyl content, and fluidity of eryth- 356. ZARLING, E. J., S. MOBARHAN, P. BOWEN, AND S. KAMATH. Pul-
rocyte membrane. Biochem. Biophys. Res. Commun. 221: 300 monary pentane excretion increases with age in healthy subjects.
303, 1996. Mech. Ageing Dev. 67: 141147, 1993.
346. YIM, M. B., J. H. KANG, H. S. YIM, H. S. KWAK, P. B. CHOCK, AND 357. ZOROV, D. B. Mitochondrial damage as a source of diseases and
E. R. STADTMAN. A gain-of-function of an amyotrophic lateral scle- aging: a strategy of how to fight these. Biochim. Biophys. Acta
rosis-associated Cu,Zn-superoxide dismutase mutant: an enhance- 1275: 1015, 1996.

/ 9j08$$ap01 03-24-98 11:38:45 pra APS-Phys Rev