Biochem. J.

(1997) 324, 1–18 (Printed in Great Britain) 1

REVIEW ARTICLE
Biochemistry and pathology of radical-mediated protein oxidation
Roger T. DEAN*§, Shanlin FU*, Roland STOCKER† and Michael J. DAVIES‡
*Cell Biology Unit, The Heart Research Institute, 145–147 Missenden Road, Camperdown, Sydney, NSW 2050, Australia, †Biochemistry Unit, The Heart Research
Institute, 145–147 Missenden Road, Camperdown, Sydney, NSW 2050, Australia, and ‡EPR Group, The Heart Research Institute, 145–147 Missenden Road,
Camperdown, Sydney, NSW 2050, Australia

Radical-mediated damage to proteins may be initiated by electron often functionally inactive and their unfolding is associated with
leakage, metal-ion-dependent reactions and autoxidation of enhanced susceptibility to proteinases. Thus cells can generally
lipids and sugars. The consequent protein oxidation is O - remove oxidized proteins by proteolysis. However, certain
#
dependent, and involves several propagating radicals, notably oxidized proteins are poorly handled by cells, and together with
alkoxyl radicals. Its products include several categories of reactive possible alterations in the rate of production of oxidized proteins,
species, and a range of stable products whose chemistry is this may contribute to the observed accumulation and damaging
currently being elucidated. Among the reactive products, protein actions of oxidized proteins during aging and in pathologies such
hydroperoxides can generate further radical fluxes on reaction as diabetes, atherosclerosis and neurodegenerative diseases. Pro-
with transition-metal ions ; protein-bound reductants (notably tein oxidation may also sometimes play controlling roles in
dopa) can reduce transition-metal ions and thereby facilitate cellular remodelling and cell growth. Proteins are also key targets
their reaction with hydroperoxides ; and aldehydes may par- in defensive cytolysis and in inflammatory self-damage. The
ticipate in Schiff-base formation and other reactions. Cells can possibility of selective protection against protein oxidation
detoxify some of the reactive species, e.g. by reducing protein (antioxidation) is raised.
hydroperoxides to unreactive hydroxides. Oxidized proteins are

INTRODUCTION (ROd) and hydroxyl (HOd) radicals. The last of these can originate
Radical-mediated protein oxidation has been studied throughout from the Fenton reaction, in which the metal ion redox cycles,
with reduction effected by O −d and oxidation by its dismutation
the century. In the first decade, Dakin (e.g. [1,2]) published #
product, hydrogen peroxide (H O ). Iron and copper are bio-
detailed chemical studies of the oxidation of leucine and other # #
amino acids by Fenton systems (transition-metal ions plus logically important transition-metal ions, with their reduced
hydrogen peroxide), and protein aggregation and (apparently) forms capable of rapidly cleaving organic (including lipid)
fragmentation were detected by others. Soon after the discovery hydroperoxides, forming radicals that can initiate chain reactions,
of glutathione, Hopkins appreciated that this reductant could be ultimately giving stable products such as lipid hydroxides.
both an anti- and a pro-oxidant, the latter depending on the Experimentally, many biologically important radicals can be
presence of transition-metal ions [3], and so could inactivate generated in defined qualities and quantities by the γ-radiolysis
proteins. Several authors assessed the proteolytic susceptibility of water. Linear fluxes of ROOd can also be generated by
of oxidized proteins, and demonstrated biphasic effects, whereby decomposition of thermolabile azo compounds such as 2,2«-
limited oxidation leads to enhanced susceptibility, while more azobis-(2-amidinopropane) hydrochloride. More complex ex-
extensive oxidation may be associated with increasing resistance perimental systems can involve the metal-ion-catalysed autoxi-
[4,5]. dation of a variety of molecules, such as sugars. In these cases,
The body of the present review focuses on the literature since it is extremely difficult to quantify the radical fluxes.
1950, and is prefaced by a synopsis of radical sources, and of the We will also refer briefly to the actions of the two-electron
chemistry of protein oxidation, to place the discussion of (non-radical) oxidant hypochlorite. This is a major product of
biochemistry and pathology in a comprehensible framework. A stimulated neutrophils, which produce superoxide radicals which
dismute to H O and then convert it into hypohalous acids by the
complementary review will detail the chemistry of protein # #
oxidation (M. J. Davies, unpublished work). action of myeloperoxidase in the presence of halides. Although
the non-radical nature of this oxidant makes it chemically
SOURCES OF RADICALS IN EXPERIMENTAL AND BIOLOGICAL distinctive, its occurrence in biological systems makes it ap-
SYSTEMS propriate for brief review.
For more detailed discussions of some of these oxidative
The primary free radical in most oxygenated biological systems systems, see [6,7].
is the superoxide radical (O −d), which is in equilibrium with its
#
protonated form, the hydroperoxyl radical (HO d). The major
#
sources of these radicals are modest leakages from the electron SYNOPSIS OF THE CHEMISTRY OF PROTEIN OXIDATION
transport chains of mitochondria, chloroplasts and endoplasmic
reticulum. Although O −d is relatively unreactive in comparison Much of the chemistry of the products of protein oxidation
#
with many other radicals, biological systems can convert it into (M. J. Davies, unpublished work) was elucidated by Dakin. He
other more reactive species, such as peroxyl (ROOd), alkoxyl detected the formation from amino acids of carbonyls, such as

Abbreviations used : apoB, apolipoprotein B ; 4-HNE, 4-hydroxynonenal ; LDL, low-density liopoprotein ; PBN, N-t-butyl-α-phenylnitrone.
§ To whom correspondence should be addressed.
2 R. T. Dean and others

Table 1 New moieties generated by biological protein oxidation and their genesis
When HOd attack is indicated, this is not intended to discriminate between the several possible HOd-generating mechanisms (metal-ion-dependent and -independent) indicated in the text or known.
Reactive nitrogen species (RNIs) similarly refers to NO, peroxynitrite, the peroxynitrite–CO2 adduct and the products of interactions of hypochlorite and hydrogen peroxide with inorganic nitrite.
Reactions of ozone and other pollutants are not considered here.

Oxidative insult Product Selected refs.

Tyr­HOd or RNIs Dopa [28]

Tyr­HOCl 3-Chlorotyrosine [223]
Tyr­RNIs 3-Nitrotyrosine [294]
Tyr­HOd, or one-electron oxidation of Tyr or HOCl, Dityrosine [134,295]
followed by radical–radical combination
Phe­HOd ; one-electron oxidation ; RNIs ? o- and m-tyrosine [295]
Phe­HOd before or after dimerization Dimers of hydroxylated aromatic amino acids [192]
Trp­HOd, or one-electron oxidation N-Formylkynurenine ; kynurenine [22,132]
Trp­HOd, or one-electron oxidation 5-Hydroxytryptophan ; 7-hydroxytryptophan [132,296,297]
His­HOd, or one-electron oxidation 2-Oxohistidine [19]
Glu­HOd in presence of O2 Glutamic acid hydroperoxide [27,31]
Leu­HOd in presence of O2 Leucine hydroperoxides and hydroxides ; α-ketoisocaproic acid ; [14,27,29,31,298]*
isovaleric acid ; isovaleraldehyde ; isovaleraldehyde oxime ;
carbonyl compounds
Val­HOd in presence of O2 Valine hydroperoxides and hydroxides ; carbonyl compounds [13,29]
Lys­HOd in presence of O2 Lysine hydroperoxides and hydroxides ; carbonyl compounds [27,31,299]†
Pro­HOd in presence of O2 Proline hydroperoxides and hydroxides ; [27,31,299,300]
5-hydroxy-2-aminovaleric acid ; carbonyl compounds
Arg­HOd in presence of O2 5-Hydroxy-2-aminovaleric acid [300]
Ile­HOd in presence of O2 Isoleucine hydroperoxides ; isoleucine hydroxides ? ; [27,31]
carbonyl compounds ?
Gly : hydrogen-atom abstraction from α-carbon Aminomalonic acid [301,302]
followed by reaction with CO2d− radicals ;
could arise from a succession of oxidations
of other amino acids
Met­HOd or one-electron oxidation Methionine sulphoxide [94,303,304]
Cys­HOd, or other hydrogen-atom-abstracting species Cystine ; oxy acids [305,306]
Non-aromatic side chains exposed to radicals ? Protein carbonyls [190]
All amino acids exposed to photo-oxidation, RCHO species formed by decarboxylation and deamination [307–309]
oxidizing radicals or HOCl
* Also S. Fu and R. T. Dean, unpublished work.
† Also B. Morin, S. Fu, M. J. Davies and R. T. Dean, unpublished work.

the oxo acids and aldehydes with the same or one less carbon believed to be important in the oxidation of side chains and the
atom than the parent amino acid ; e.g. glycine giving rise to backbone respectively.
glyoxal and glyoxylic acid, formaldehyde and formic acid ; alanine A more recent advance has been the realization that, in the
giving rise to acetaldehyde and acetic acid, etc. This general presence of O , there is a protein-radical chain reaction, with
#
scheme has been confirmed for many amino acids, including modest O consumption and a chain length of seven or more
#
aromatic amino acids, although other reactions may also occur. [21,22], as indicated but not identified in earlier data [23,24]. The
Notable quantitative studies were contributed by Garrison, who possible participation of a range of radicals in this chain process
also proposed a ‘ peptide α-amidation ’ scheme for the oxidative has been demonstrated by detailed EPR spin-trapping studies
breakage of polypeptide backbones ([8,9] ; reviewed in [10]), [25,26]. Alkoxyl radicals apparently have a greater importance in
which has been confirmed in some cases [11], and seems to be a protein oxidation chains [25] than they do in lipid peroxidation,
common component of chain-fragmentation reactions. in which peroxyl radicals are the key chain-propagating species
During the oxidation of aliphatic amino acids by HOd, [6].
hydroxylated derivatives, notably of the side chains, are formed. Two categories of reactive, but non-radical, intermediates in
These were partially characterized by Kopoldova and co-workers protein oxidation have been identified. Reductive moieties,
(e.g. [12]), and have been fully defined for valine [13] and leucine notably dopa formed from tyrosine, can reduce transition-metal
[14]. During the oxidation of aromatic residues, the formation of ions, thus enhancing reactions with hydroperoxides, and are also
phenoxyl radicals from tyrosine, and their conversion into able to induce radical formation in reactions with O [27,28]. The
#
dityrosine and further products, can occur, especially if there are other category is the hydroperoxides formed particularly on
no reductants to repair the tyrosyl radicals (e.g. thiols, vitamin E) aliphatic side chains, but probably also on main-chain α-carbons
and if there are vicinal tyrosyl radicals [15,16]. Hydroxylation of [14,26,27]. These can be decomposed by transition-metal ions to
phenylalanine, tyrosine and tryptophan is also a characteristic give further radicals, which may propagate reaction chains. The
reaction of hydroxyl radicals [17,18], and similar reactions of hydroperoxides may also be reductively detoxified to hydroxides,
histidine (giving 2-oxohistidine) are important [19]. Fenton probably without radical formation [29]. Hydroperoxides on
chemistry can generate both the aliphatic and aromatic [20] proteins were first detected in the 1940s [30], but were only
products. Table 1 summarizes some products of protein oxi- chemically characterized more recently [27,31].
dation ; Schemes 1 and 2 illustrate some of the major reactions The precise stoichiometry of the protein oxidation chain is
 Radical-mediated protein oxidation 3

~NH CH C(O)~

.
X

Further reactions
~NH CH C(O)~

.
C ~NH CH C(O)~
. Side-chain radical
R R
Side-chain alcohols/hydroxides

O2
RH
~NH CH C(O)~
~NH CH C(O)~ ~NH CH C(O)~
2× C OOH
H+
. .
C O Via tetroxide, which C OO
Side-chain hydroperoxide
decomposes
to form two alkoxyl radicals Side-chain One-electron oxidation
peroxyl radical
Side-chain alkoxyl radical
One-electron reduction

Rearrangement to give carbon- . Side-chain alkoxyl radical

R + Carbonyl compound
centred radical via
1,2 shift reaction
Carbon-centred radical and carbonyl compound
from β-scission of tertiary alkoxyl radicals

α-Hydroxyalkyl radical
~NH CH C(O)~
.
~NH C C(O)~
Oxidation
R R
O2

α-Carbon backbone radical

α-Hydroxyalkyl peroxyl radical Carbonyl compound

.
HOO

Further reactions

Further reactions

Scheme 1 Major reactions of aliphatic side-chain radicals formed during protein oxidation in the presence of oxygen
Species detected as products of side-chain oxidation are depicted in colour.

unknown. It is clear that there are several propagation reactions other enzymes were carried out mainly in the absence of O . They
#
that do not require O , consistent with the low O consumption showed that HOd was the most effective inactivator, and charac-
# #
observed, but the overall reaction is O -dependent, probably as terized other more selective (but less efficiently inactivating)
#
a result of a requirement to form peroxyl (and possibly alkoxyl) species such as (SCN) −d, Br −d, Cl −d and I −d. For example,
# # # #
radicals from carbon-centred species. (SCN) −d was found to react with an important tryptophan
#
residue in pepsin and so inactivate the enzyme, although damage
could be reversed by the same radical [34]. Inactivation by the
BIOCHEMISTRY OF PROTEIN OXIDATION hydrated electron has also been reported (e.g. [35]), but its
Reversible and irreversible inactivation of proteins by radicals significance, and that of the above-mentioned selective radicals,
for biological systems may be limited. In studies on -amino acid
Damage by radiation-induced radicals oxidase, it was found that removal of the coenzyme FAD
Early radiation studies on lysozyme [32], ribonuclease [33] and enhanced radical damage and inactivation, illustrating that
4 R. T. Dean and others

~NH CH C(O)~

.
X

.
~NH C C(O)~
~NH CH C(O)~
R
R
α-Carbon radical

O2

.
. OO OOH
O
2× ~NH C C(O)~ ~NH C C(O)~
~NH C C(O)~ H+
Via tetroxide, which
R decomposes R R
to give two alkoxyl radicals One-electron oxidation
α-Carbon hydroperoxide
α-Carbon
peroxyl radical
RH One-electron reduction
Further reactions
.
R . α-Carbon alkoxyl radical
HOO

OH
H2O ~NH CH C(O)~
~NH C C(O)~ ~NH C C(O)~
R
R R
α-Carbon alcohol
~NH CH C(O)~

Backbone imine .
C
β-Scission
Hydrolysis

O Side-chain carbon radical

~NH2 + RC(O)-C(O)~
.
~NH C + C(O)~

Further reactions

Further reactions

Scheme 2 Major reactions of backbone radicals formed during protein oxidation in presence of oxygen
Species detected as products of backbone oxidation are depicted in colour.

conformation [36] and ligands can affect the extent of inactivation These studies developed from earlier work on the oxidative
(see [6] for a review of inactivation studies). inactivation of glutamine phosphoribosylpyrophosphate amido-
transferase in extracts from Bacillus subtilis [37,38]. This system
is O -dependent and is perturbed by metal-ion chelators, but its
Damage by metal-ion-catalysed systems #
chemistry has not been fully defined. A similar inactivation
Since 1981, Stadtman and colleagues have examined the in- occurs in a mutant strain lacking important proteinases ; oxi-
activation of proteins in cell-free systems involving the metal- dation of a (4Fe–4S) centre appears to be important.
ion-catalysed autoxidation of ascorbate and}or hydrogen per- Although metal-catalysed oxidation systems are mechanisti-
oxide ; in some cases the metal ions were derived from metallo- cally diverse, they are like Fenton systems and involve H O ;
# #
proteins. These systems are now termed ‘ metal-catalysed oxi- thus they can be blocked by catalase. They inactivate a wide
dation systems ’, instead of the previous confusing term ‘ mixed- range of enzymes (e.g. [39–41]) and are suggested to be important
function oxidation systems ’. for both proteolytic turnover [39] and the accumulation of
 Radical-mediated protein oxidation 5

proteins during aging [42]. The inactivation of one of the most inhibitor (e.g. [55]) ; they also inactivate proteinase inhibitors and
studied enzymes, glutamine synthetase, is influenced by its neutrophil elastase, which lack the susceptible methionine, with
adenylation state [39], which also regulates the enzyme and some comparable efficiency [56]. Thus methionine-containing inhibitors
multi-enzyme cascades [43,44]. are not automatically more vulnerable, and susceptibility depends
Some metal-catalysed oxidation systems can cause selective on the radical or oxidizing conditions. Which systems are
damage in cell-free systems, such as to the histidine residues in important in ŠiŠo are not known.
glutamate synthetase (reviewed in [45]), but this is accompanied
by other alterations ; in this case, both protein fragmentation [46]
and changes in hydrophobicity. Limited oxidation increases Damage by singlet oxygen and photochemical reactions
hydrophilicity, while further oxidation increases hydrophobicity Singlet oxygen ("O ) and radical species can participate in
[47]. Early studies suggested that modification of one (of 16) #
photosensitizer-induced inactivation of enzymes, as exemplified
histidines both inactivated and increased proteolytic suscep- by studies on purified catalase, or catalase within cells [57].
tibility [48] ; later studies showed that enhanced proteolytic Photosensitized inactivation of protein kinase C due to calphostin
susceptibility requires the modification of at least two histidines [58] has been shown to be reversible and dependent on O .
per subunit [49]. By the stage at which this takes place, other #
Within cells this process becomes irreversible, and oxidation
modifications also occur, e.g. at Arg-344 [50], yet only 0.7 mol of appears to occur mainly on the membrane portion of the complex.
carbonyl}mol of protein is present [49], so most altered histidine
residues do not contain carbonyls. Incorporation of tritium
(from tritiated borohydride) into the enzyme at various stages of Damage by endogenous radicals
oxidation revealed eight (unidentified) radioactive peaks using
amino acid HPLC after protein hydrolysis. Some of these may It is known that a number of enzymes contain an intrinsic radical
represent carbonyls. As the chromatographic pattern was rela- species while in their active form (reviewed in [59]). These
tively constant at different times, it was argued that the range of radicals can be present on tyrosine, modified tyrosine, tryp-
products is also constant. However, during the period when the tophan, modified tryptophan, glycine or thiols. For example,
first histidine was oxidized (! 60 min of oxidation), less than six during catalysis ribonucleotide reductase contains a tyrosyl
of these species were formed. By 120 min little further histidine radical, and probably also a thiyl radical, in order to abstract a
oxidation had occurred, yet a further two or three amino acid hydrogen atom from the ribose ring of the substrate [60].
The reaction of these protein radicals with O (or possibly
derivatives were formed. Unfortunately, the earliest time point #
reported was 20 min, making it difficult to conclude that the other species) can cause enzyme inactivation by backbone
critical event for inactivation is the histidine oxidation, since cleavage rather than side-chain alterations. Reduction of the
other modifications had also occurred. Consumption of other intrinsic tyrosyl radical on the R2 subunit of the enzyme can be
individual amino acids was low when compared with that of accompanied by cleavage of the R1 subunit into 26 and 61 kDa
histidine, and could not be detected ; the aggregate loss of all fragments, respectively from the N- and C-termini of R2. The
amino acids was not quantified, and may have been significant. precise site and mechanism of this reaction [61] remain to be
In ŠiŠo studies of glutamine synthetase inactivation have been elucidated. Similarly, inactivation of anaerobically grown
carried out in other species, such as the fungus Neurospora Escherichia coli pyruvate formate lyase, which contains an
crassa, where the inactivated enzyme has been detected in ŠiŠo intrinsic glycyl radical, can occur in the presence of O or
#
[51]. Furthermore, when living Klebsiella pneumoniae [52] is hypophosphite and appears to involve glycyl radical reactions.
switched from anaerobic to aerobic conditions (or exposed to This process is initially reversible, since damage is both localized
and limited in extent [62]. Reversal may involve loss of O from
H O ), several enzyme activities, including that of glycerol
# # #
dehydrogenase, are lost, apparently by oxidative inactivation. ROOd, a relatively facile process when the resultant carbon-
Purification of this enzyme when 90 % inactivated (from cells centred species is stabilized (as it will be here, since backbone α-
exposed to H O ) and comparison with the native form has carbon radicals are capto-dative species stabilized by both the
# # carbonyl and amide functions). This glycyl α-carbon radical
shown no detectable differences in either subunit molecular size
or amino acid composition (including -SH groups). The lack of occurs in the sequence -Ser-Gly-Tyr- which is cleaved with
such gross changes suggests that limited modifications are formation of an N-terminal oxalyl residue from the glycine [11],
sufficient for inactivation, although the inactivated enzyme was consistent with the α-amidation fragmentation mechanism [10].
more hydrophobic, and its intact oligomer migrated in gel Many oxidative enzymes, such as lipoxygenases and the
filtration as if slightly larger than the native form. This emphasizes cytochrome P-450 family, can also generate radical species during
a major problem in studying oxidative inactivation in ŠiŠo : even their interaction with substrates, such that ‘ self-inactivation ’
analysis of the inactivated enzyme from these cells gave no occurs. Thus lipid peroxyl radicals are released during lipoxy-
evidence that inactivation was due to protein oxidation, although genase action, and the product hydroperoxides of linoleic acid
it was clearly initiated by an oxidative affront. can inactivate it when iron is available (from the enzyme or
elsewhere). Inactivation is enhanced by O , and such reactions
Inactivation of non-enzymic proteins by metal-catalysed oxi- #
dation has also been studied extensively. Thus α -proteinase give rise to the ‘ hydroperoxidase ’ activity of this enzyme [63].
"
inhibitor is inactivated on oxidation of a susceptible methionine
residue, and H O inactivates a neutrophil cytosolic serine-
# # Protein damage by radicals or carbonyls arising from other biological
proteinase inhibitor (serpin), possibly via a similar process [53].
In some cases, these reactions of H O and methionine may be
processes
# #
nucleophilic (molecular) rather than radical-mediated. The pro- Radicals produced during lipid autoxidation can inactivate α -
"
teinase subtilisin is dependent on a methionine residue both for proteinase inhibitor by oxidation of Met-358 [64]. Similarly, the
its activity and for its susceptibility to H O in Šitro ; replacement formation of haemichrome is concomitant with lipid oxidation
# #
of this residue by site-directed mutagenesis decreases inactivation during Fe(II)-induced oxidation of liver slices [65]. Later studies
of the enzyme by H O [54]. In contrast, Fenton systems produce with liver homogenates also demonstrated a slow inactivation of
# #
a range of amino acid derivatives that inactivate α -proteinase glutathione peroxidase [66].
"
6 R. T. Dean and others

End-products of lipid oxidation, such as malondialdehyde [67] aromatic residues are also oxidized [87], suggesting a radical-
and 4-hydroxynonenal (4-HNE) [68], are also inactivating agents, mediated process. A number of neutrophil neutral proteinase
possibly via Schiff-base formation. Schiff bases are short-lived inhibitors, isolated E. coli glutamine synthetase and BSA when
species formed by the reaction of carbonyl groups with amines, exposed to ozone also show methionine oxidation and lesser
and can be formed, for example, during exposure of proteins to : oxidation of histidine and aromatic residues [83,88]. The formyl-
(i) lipid-derived aldehydes [69], (ii) autoxidizing sugars (some of Met-Leu-Phe-triggered respiratory burst of neutrophils also
which form vicinal dicarbonyl compounds [70]), and (iii) amino results in oxidation of the methionine in the formyl-Met-Leu-Phe
acid-derived aldehydes, as mentioned earlier. Thus binding of the [89].
apolipoprotein B (apoB) protein of low-density lipoprotein Two pathways may exist for the reversal of methionine
(LDL) to its cellular receptor is perturbed by reaction with 4- oxidation. Firstly, the initial radical species derived from meth-
HNE at modest levels [71], and grossly altered when aggregating, ionine (i.e. the radical cation or the corresponding hydroxylated
supra-pathological, levels are used [72]. With some lipid material) can be reduced by suitable donors such as the water-
carbonyls, Schiff-base formation may be of limited importance soluble tocopherol analogue Trolox C [90]. Secondly, methionine
compared with Michael addition reactions [73,74]. In insulin sulphoxide can be reduced by methionine sulphoxide reductase,
(which lacks cysteine), histidine residues are selectively modified which is present in many mammalian cells [91,92]. The latter may
by Michael addition of 4-HNE [75]. With dialdehydes, Schiff- be the major re-activating process with α -proteinase inhibitor
"
base formation and Michael addition appear to be simultaneous, [64]. Further oxidation products of methionine sulphoxide (e.g.
thus cross-linking lysine residues [76]. 4-HNE also inactivates the sulphone [93]) cannot be reversed by the reductase. Meth-
glucose-6-phosphate dehydrogenase and glyceraldehyde-3-phos- ionine oxidation and repair may be a biological control mech-
phate dehydrogenase ; in the latter case both intra- and inter- anism [94], analogous to reversible thiolation.
molecular cross-links appear to be formed [77].
Schiff bases
Mechanisms of reversal of oxidative lesions Schiff-base adducts formed on reaction of an amine with a
Thiol groups carbonyl are unstable [95], and so may be reversible inactivating
lesions. Schiff-base formation is, however, often followed rapidly
Some of the (limited) modifications discussed above are tran- by Amadori rearrangements (e.g. [96]), and hence the window for
siently reversible, and such reversal may cause re-activation. possible reversal is small. The thioether linkage formed by
Conversion of -SH groups into disulphides and other oxidized Michael reactions of protein -SH groups with 4-HNE is unlikely
species (e.g. oxyacids) is one of the earliest observable events to be reversed biologically. During hypochlorite oxidation of
during the radical-mediated oxidation of proteins, with the ratio LDL, in which protein is the preferential target, lysine-derived
of disulphide to oxygenated materials depending on the en- aldehydes seem to form protein cross-links that are initially
vironment. For example, H O causes preferential thiol (and reversible [97].
# #
subsequently methionine) consumption in lens crystallins in Šitro,
whereas most other amino acids show little change ; this may be Reversible and irreversible protein unfolding
a non-radical (molecular) reaction [78]. Higher sulphides and
persulphides may also be generated on proteins [79], resulting in Very limited protein unfolding occurs in most of the examples of
conformational changes [78]. (cell-free) protein inactivation given above. Similarly, in ŠiŠo
Inactivation of enzymes by limited -SH oxidation has been alterations to glycerol dehydrogenase during exposure of
studied, mainly with oxidized thiols (e.g. GSSG) as oxidants, so Klebsiella to H O [52] involve minor size and hydrophobicity
# #
as to restrict the range of reactions. Thus treatment of aldose changes. This is in agreement with the small free-energy changes
reductase with GSSG results in mixed disulphide formation, (0.25–0.5 kJ}mol) measured during the in Šitro inactivation of
conformational changes and inactivation of the enzyme, which glutamine synthetase [98].
can be reversed by GSH [80]. Reversible S-thiolation has also In Šitro studies also show that alteration of protein size occurs
been detected on proteins in cells exposed to radicals [81]. to only a limited extent during oxidation. Thus BSA exposed to
Furthermore, the respiratory burst of human monocytes results HOd (from radiolysis) in the presence of O yields only small
#
in rapid and reversible S-thiolation of a number of cytosolic amounts of molecules of changed size, as detected by non-
proteins [82]. Thus biological thiols, such as GSH and cysteine, denaturing HPLC gel filtration, even though dose-dependent
can influence oxidant-induced protein inactivation either by generation of smaller fragments can be seen by SDS}PAGE
direct reaction with the radicals or indirectly, by forming [99,100]. In the absence of O , HOd causes substantial cross-
#
reversible bonds with (normally free) thiols on the proteins. The linking. Similar observations have been made during prolonged
latter should be reversible (e.g. [80,82]) ; it is also an essential in Šitro autoxidation of lens crystallins [101] ; later studies
process in de noŠo protein folding and in the maintenance of identified aggregates by light scattering and size-exclusion
conformation. chromatography [102]. Thus oxidatively generated fragments
probably remain associated, at least initially, in a conformation
similar to that of the native protein.
Methionine As mentioned above, inactivation often involves limited con-
The oxidation of methionine residues, discussed above for formational change. On the other hand, limited in Šitro oxidation
proteinase inhibitors [83,84], can be reversible. t-Butyl hydro- of glyceraldehyde-3-phosphate dehydrogenase and alcohol de-
peroxide behaves like H O , but only oxidizes exposed hydrogenase by "O or HOd causes changes to the intrinsic
# # #
methionines [85], presumably because its increased steric bulk fluorescence [103]. This raises the question as to whether limited
limits access. Treatment of isolated human erythrocyte conformational changes (or even large-scale unfolding) can be
glyceraldehyde-3-phosphate dehydrogenase and glycophorin reversed, perhaps independently of the repair of the chemical
with ozone also leads to the loss of both methionines. In the lesions responsible for them. Very limited unfolding can be
latter, no other amino acids are oxidized, and the reactions are reversed (e.g. [103]), but there have been no systematic studies
probably nucleophilic [86]. In the former, however, cysteine and with radical-damaged proteins.
 Radical-mediated protein oxidation 7

Protein disulphide isomerase [104] is concerned with re-pairing an initial loss of ε-amino groups of lysine, and a subsequent
of -SH groups, while chaperonins are important in controlling increase in soluble amino groups ; if the lysine groups were
correct folding after protein synthesis [105]. In theory, the various blocked prior to oxidation, only modest increases were observed
conformations accessible to a given polypeptide are exchange- [112]. Thus, until precise measurements of the competing path-
able, and these proteins can facilitate the exchange, but the ways are obtained, the release of trichloroacetic acid-soluble
energy barriers to such interconversions may be too great for amino groups is an unreliable index of fragmentation. Accurate
them to occur at a detectable rate under physiological conditions. quantification of released materials requires knowledge of the
The likelihood of the correct refolding of a substantially damaged exact amounts of protein present. Determination of Kjeldahl
protein is probably low, even if it is recognized by chaperonins. nitrogen or total amino acids after hydrolysis provides such data,
The situation during oxidative insult may, however, be although this can be difficult to achieve with oxidized fragments.
different. It is possible that chaperonins among the heat-shock A more reliable method of measuring fragmentation is to use
protein families that are induced by oxidative stress may protect proteins radioactively labelled in either main- or side-chain sites.
against the irreversible denaturation of a partially unfolded These can be produced by biosynthesis, reductive methylation
molecule. Thus macrophage colony-stimulating factor, which [110] or iodination. With the last of these, routinely used in
enhances O −d output, induces the synthesis of heat-shock proteins cellular metabolism studies, the free iodide generated can be
#
of 60, 70 and 90 kDa, and confers enhanced resistance to H O ; problematical and needs to be removed by treatment with silver
# #
the chaperonins may contribute to this effect [106] and protect nitrate.
against autoxidative damage during the respiratory burst. A
parallel induction of chaperonins arises on exposure of the
intracellular facultative bacterium Francisella tularensis LVS to
Features of radical-mediated protein fragmentation and polymerization
H O ; this is presumably a defence against oxidative stress A series of important studies [109,113–115] has shown that, even
# #
produced by the host macrophages [107]. Many human cells also though HOd attack on BSA causes extensive O -dependent
#
respond to oxidative stress by the induction of heat-shock fragmentation, this process can be selective. Cross-linking (both
proteins (e.g. [108]). Complexes of partially unfolded proteins reductant-sensitive and -resistant links) is also observed, and
with chaperones may trigger further responses in these systems. predominates in the absence of O . Fragment formation can be
#
The lens of the eye is especially vulnerable to oxidative damage, detected during anoxic irradiation, at least when lactate de-
as a result of both high exposure levels to UV light and the fact hydrogenase is the target [116]. These observations have been
that the crystallin proteins are exceptionally long-lived. Thus it is confirmed and extended [24,99,100,110,117]. HOd and O −d act
#
not surprising that α-crystallin is a chaperone that can restrict the synergistically [99], possibly as a result of additional decay
aggregation of other proteins, and hence opacity and cataract pathways for the α-carbon peroxyl radicals by reaction with
formation [102]. Oxidation of this crystallin in Šitro diminishes its O −d}HOOd The reasons for selective cleavage are not entirely
#
chaperone capacity, and α-crystallin isolated from senile human clear ; they may include differences in accessibility of sites to the
lenses shows such decreased chaperone activity and enhanced attacking species, the stability of the protein radicals formed, and
oxidation. the occurrence of radical transfer reactions.
Enhanced susceptibility to degradation by proteinases is often The sizes of the fragments generated from BSA, haemoglobin
employed as a criterion of unfolding. This criterion is usually, and myoglobin by radiolytic HOd are compatible with frag-
but not always, fulfilled with oxidized proteins. Thus alcohol mentation at proline residues [109,115]. Indeed proline (and
dehydrogenase is relatively more resistant to elastase and pro- histidine) is an important site of HOd attack on BSA, although
teinase K than several other proteins after limited oxidation by the predicted product(s) from cleavage at this site could not be
"O or HOd [103]. It is known that extensively oxidized proteins detected, while new products arising from proline were formed
#
may be denatured and aggregated to such an extent that [118]. This putative pathway has been observed in simple peptide
proteolytic access is restricted rather than enhanced ; the dis- systems [119], but is inconsistent with the reported stability of the
tinction between limited and extensive is, of course, arbitrary and α-carbon radical formed from proline (see [120]).
wholly dependent on the protein under study. The other major We have re-assessed the fragment size data obtained by
component of irreversible protein oxidation is fragmentation, Schuessler and co-workers from BSA, haemoglobin and myo-
and when this is extensive it seems inevitably to lead to loss of globin [109,114,115], and conclude that they can be explained
conformation. more readily by cleavage at glycine (M. J. Davies, unpublished
work), whose α-carbon radical is especially stable [121]. Glycine
Protein fragmentation and polymerization residues also appear to be important in the fragmentation of calf
skin collagen induced by a xanthine oxidase system [122].
Measurement Cleavage is probably induced by HOd in this system, as the
The measurement of oxidative protein fragmentation, which primary O −d produced is inactive in chain breakage. Of the N-
#
does not always involve the formation of new N-termini, is not terminal amino acids generated by fragmentation, 90 % were
straightforward. In most cases only qualitative measurements glycine, even though the parent molecule contains only approx.
have been made. Quantification by SDS}PAGE under reducing 30 % glycine. Selective fragmentation has also been observed
conditions (e.g. [109]) is difficult, as the fragments are often small with the apoB protein of LDL during either radiolysis or metal-
and difficult to retain (during electrophoresis and}or staining), ion-catalysed damage [123] ; it is difficult to to assess the relevance
and individual fragments may stain differentially depending on of cleavage at proline or glycine in this case, due to difficulties in
their composition and degree of oxidation [110]. measuring fragment sizes. The importance of glycine residues
There have been attempts to correlate fragmentation with during radical attack on proteins in aqueous physiological
terminal amino group generation. When amino groups were systems (cf. studies in organic solvents [124,125]) will only be
measured in proteins exposed to HOd after unfolding induced by resolved by sequencing of the fragments.
guanidine hydrochloride, modest increments in total amino Comparison of the data from metal-ion}peroxide with those
groups were observed after low doses, with losses at higher doses from radiolytic systems shows that the fragmentation patterns
[111]. Similarly, exposure of lysozyme to Cu(II)}H O resulted in are distinct ; this has been confirmed by direct comparison
# #
8 R. T. Dean and others

[126–128]. The selective nature of metal-ion-catalysed oxidation His, Cys, Lys and Met, which bind metal ions, may localize
may be a consequence of the localization of the metal ions at reactions to their vicinity. The differential capacities of proteins
particular sites on the protein [129–131]. to bind metal ions and render them either redox-inactive (e.g.
For some proteins (e.g. haemoglobin), unlike BSA, the extent transferrin, lactoferrin) or active also influences the distribution
of HOd-induced fragmentation is similar with or without O . This of damage among protein populations. Few studies have directly
#
may arise as a result of the presence of the iron–protoporphyrin addressed the quantitative differences in distribution of damage
IX prosthetic group, which may liberate sufficient O from H O in heterogeneous systems.
# # #
generated during radiolysis to permit fragmentation and restrict
cross-linking. A similar rationale may explain the absence of
cross-linked material from BSA treated with Cu(II)}H O [126].
# # Protein oxidation by lipid-derived species
Both aggregation and fragmentation have been observed in a
similar system with lysozyme [112], but the aggregation is It is well established that lipid radicals may damage proteins (e.g.
coincident in time with the loss of lysine amino groups, and may [141] ; reviewed in [142]). Such reactions show differences from
be non-covalent. those observed with radiolytic systems, as exemplified by the
Using proteins that lack one of Tyr, Trp or His, Guptasarma protein fragmentation patterns observed with model aqueous}
et al. [132] demonstrated the importance of His in covalent multiphasic systems [127]. In systems that contain lipid hydro-
protein cross-linking in the presence of O . Lys was also peroxides and soluble proteins, metal-ion-catalysed reactions
#
demonstrated to be involved ; these cross-links may involve appear to be central. Thus amino acid consumption varies
Schiff-base derivatives. In the Tyr-containing proteins bovine somewhat between proteins, as might be expected if site specificity
pancreatic trypsin inhibitor, RNase A and crystallins, dityrosine plays a role [143]. Furthermore, protein inactivation by
cross-links were not observed. Interestingly, these workers also peroxidizing lipid is often associated with the binding of lipid
noted that treatment of melittin (which lacks Tyr) with HOd components to the protein. Thus, in membranes, competition
induced changes in protein fluorescence characteristic of the and interactions between protein and lipid oxidation are expected.
formation of dityrosine, demonstrating the danger of identifica- Early studies emphasized the fluorescent cross-links that can
tion of dityrosine on this basis. Definitive identification can be form between lipid oxidation products and proteins, and their
obtained by the use of HPLC separation combined with mass possible contribution to ceroid, lipofuscin and other ‘ age
spectral identification (e.g. [133]) and co-elution. This has been pigments ’ found in cells. The demonstration of such fluorophores
done in some (e.g. myeloperoxidase-catalysed oxidation of in Šitro has not generally been followed by the precise chemical
proteins and exogenous Tyr [134]), but not all [135], studies. definition of the materials in ŠiŠo (reviewed in [143]).
Aggregation of fibrinogen as a result of treatment with The interactions between lipids and proteins during protein
Cu(II)}ascorbate is probably a special case, since chain cleavage oxidation in membranes have been studied using the mito-
may lead to aggregation [136], as it does during proteolytic chondrial outer membrane enzyme monoamine oxidase, selec-
conversion into fibrin. However, oxidized fibrinogen is not tively labelled (covalently) with a radioactive binding inhibitor.
cleaved normally by thrombin [137], in agreement with early Monoamine oxidase can be delipidated subsequently to labelling.
irradiation studies [138]. Fibrinogen appears to be the plasma When present, lipid consumes some of a radiolytic radical flux,
protein that is most sensitive to in Šitro metal-ion-catalysed such that the extent of monoamine oxidase fragmentation is
carbonyl formation [139]. decreased. The fragmentation yield in these circumstances is also
lower than that with BSA. However, after initial radiolytic
oxidation, membrane-bound monoamine oxidase is much more
Protein oxidation in solution, in membranes and in lipoproteins susceptible than soluble BSA to further damage by metal ions.
The immediate environment of polypeptides influences the nature This is apparently due to radicals derived from the lipid hydro-
and extent of their reactions with radicals. The key factors are peroxides formed in the initial oxidation [101,144–146]. Metal-
the protein concentration (and if there is more than one, the ion-catalysed damage to unirradiated membranes seems to affect
nature and relative concentrations), and the nature, extent and lipid and protein in parallel, and is restricted by α-tocopherol
location of the radical flux. The yield of oxidation products for [145]. Membranes can also influence radical damage to proteins
a given radical flux increases with protein concentration until the by physical separation of the target from the radical source [147]
system is saturated (i.e. all incident radicals react with the protein or by sequestration of transition-metal ions away from hydro-
rather than with other components). Under saturating conditions peroxides [148].
and with non-selective oxidants, different proteins would be Lipoproteins are specialized lipid–protein environments, and
expected to interact with the incident radicals in a manner their relevance to atherosclerosis (see below) has lead to significant
dependent on their relative volume occupancy. Whether this work on their oxidation. Lipid oxidation products, notably
results in proportionate damage depends on whether inter- aldehydes, can modify the lysine residues of apoB, and thereby
protein radical transfer occurs and whether the oxidation chains decrease binding to the classical LDL receptor, while increasing
on each protein are of comparable length. its endocytosis by other routes, such as the scavenger receptor.
Intra-peptide and -protein transfer of radicals has been LDL oxidation by metal-catalysed oxidation causes generalized
demonstrated directly (reviewed in [6,140]). Indirect evidence loss of amino acids, although aspartate and glutamate are
also supports the occurrence of such reactions ; thus aromatic increased [149], consistently with the damage to histidine and
residues are normally damaged more extensively than aliphatic proline residues. Fragmentation also takes place, as demonstrated
residues (although all are usually damaged to a significant by selective radiolytic techniques [123]. During metal-ion and
degree), despite the fact that many of these residues are internal radiolytic attack, lipid and protein oxidation are often con-
to a protein structure and hence not readily accessible to radicals current, and occur even while much vitamin E remains in the
in solution. Inter-chain interactions are a prerequisite for cross- particles. In contrast, hypochlorite selectively attacks the protein,
linking reactions. consuming mainly lysine, tryptophan, cysteine and methionine
With metal-ion-catalysed oxidations, the location of the active residues, and giving rise to chloramines [97,150]. Exogenous
transition-metal ion becomes important, and residues such as tyrosine can be cross-linked to apoB tyrosines by hypochlorite
 Radical-mediated protein oxidation 9

[150], although it is not clear whether this occurs with physio- proteolytic removal in living cells. Alteration in protein -SH}S–S
logical levels of tyrosine. status was one of the first signals proposed [172,173]. Developing
earlier work [37,38] Stadtman and colleagues [39] integrated
metal-catalysed oxidation with initiation of proteolysis. They
Protein oxidation in isolated organelles and other complex systems suggested that limited modifications (e.g. to one histidine residue
Studies on phage inactivation by radicals, and its sensitization by in glutamine synthetase) might suffice to signal degradation.
O and localization by transition-metal ions, are indicative of However, even in simple lysate systems this proposed relationship
#
both protein and DNA damage (e.g. [151]), although protein between susceptibility to degradation and single modifications to
damage and DNA–protein cross-linking were directly studied the target is difficult to verify, since many alterations occur
only to a limited extent [152]. Damage to mitochondrial electron- concomitantly (see above ; [49]). These difficulties are multiplied
transport proteins by hypochlorite [153] and other agents has when modifications within living cells are considered. Even in the
been studied, and noted to be enhanced by vitamin E deficiency most carefully studied systems [52,174–177], one can only rig-
[154]. Mitochondria exposed to exogenous radicals lose control orously conclude that a wide range of oxidative modifications are
of ion balance, notably of calcium transport ; protein oxidation concomitant with elevated proteolytic removal. Similar observa-
as well as proteolysis may be important in such changes tions have been made with isolated mitochondria and choloro-
[145,155,156]. Leakage of electrons from the chains, leading to plasts, so that the idea probably has generality.
radical fluxes and self-inactivation, may also be important, Within cells there are many potential proteolytic sites (mito-
especially as mitochondria seem to be a major radical-generating chondria, lysosomes, proteasomes) for degradation of oxidized
site and contain more oxidized DNA than nuclei [157,158]. proteins, some of which originate locally, others of which are
Again, this may be associated with alterations in enzymic transported there. Selective transport of certain cytosolic and
proteolysis as well as protein oxidation [159,160]. These studies organelle proteins to lysosomes for degradation appears to be
indicate close parallels between lipid and protein oxidation in important [178–180] ; in contrast, selective reverse translocation
mitochondria. of newly synthesized unfolded proteins from the endoplasmic
In chloroplasts, electron-transport proteins are continuously reticulum to the proteasome can occur [181].
exposed to radicals derived from electron leakage and photo- The endocytic supply of oxidized protein substrates is one of
chemical reactions. This is important, for example, in the few approaches in which one can define the quantities of
inactivation of photosystem II, in which a 32 kDa herbicide- substrates, their catabolic rate and their locations within the cell.
binding protein is oxidatively modified prior to proteolytic ApoB of oxidized LDL follows a normal transport pathway
degradation [161,162]. within macrophages, but is poorly degraded and accumulates (in
Microsomal oxidation reactions depend largely on the cyto- lysosomes) to an elevated level compared with normal LDL
chrome P-450 family, and these enzymes seem to undergo suicidal apoB [171,182]. The precise importance of apoB itself is not
inactivation, independently of lipid oxidation. In this case the clear, since oxidized lipids are also supplied. However, 4-HNE-
substrate probably forms radicals in, or near, the enzyme active derivatized apoB in otherwise unaltered LDL, as well as oxidized
site, leading to haem damage and inactivation [163,164]. (lipid-free) BSA, are also poorly degraded, suggesting that protein
Ascorbate can protect microsomes against protein damage modifications are important [171,183].
involving cytochrome P-450, which seems to be a conventional For intracellularly generated oxidized proteins, the degra-
metal-catalysed oxidation system, using the iron of the cyto- dation sites may be multiple, and require identification. Evidence
chrome [165,166]. for a crucial role for the proteasome in the degradation of such
In these membranous systems, therefore, there may be both proteins is unconvincing [184]. The support from studies using
lipid-dependent and -independent pathways of protein oxidation, antisense RNA inhibition of proteasome function [185] appears
just as there may be protein-independent lipid oxidation. In to be undermined by gross re-utilization of radioactive amino
contrast, in extracellular connective tissue, protein oxidation acids, such that the measured ‘ degradation ’ rates are implausible
appears to be an isolated process, removed from lipid, although and more probably reflect protein synthesis rates.
occurring in the vicinity of high concentrations of proteoglycans. Further work is needed to compare the absolute rates of
During radiolytic or Fenton-derived radical attack on cartilage catabolism and the half-lives of oxidized and native proteins.
discs, polypeptides are preferentially attacked by HOd [167,168], This requires knowledge of both pool sizes and catabolic rates.
although these radicals can also depolymerize the non-pro- Microinjection studies have shown that the proportional degra-
teinaceous components [169]. dation of oxidized proteins can be more rapid than that of the
parent [186]. However, difficulties remain in extrapolating data
from these necessarily artificial conditions (microinjection, in-
Enzymic removal of oxidized proteins appropriate pool sizes) to normal conditions. One promising
Like most partially denatured proteins, modestly oxidized approach may be to use purified oxidized amino acids as
proteins are usually more sensitive to proteolytic attack by most substrates for protein synthesis, thereby generating controlled
proteinases [39,99,117], whereas heavily oxidized proteins gen- amounts of ‘ oxidized ’ proteins without oxidative affront to the
erally show decreased susceptibility [101,117]. The latter may cells. A detailed review of the proteolysis of oxidized proteins is
depend on local chemical and structural changes rendering parts available [187].
of the molecules poorly digestible [170], in agreement with the If the degradation of oxidized proteins in ŠiŠo is efficient, one
fact that certain aldehyde derivatives of proteins can be less would expect a small but finite pool of oxidized amino acids to
susceptible to proteolysis than the parent protein ; for example 4- be present in normal proteins. Table 1 indicates this to be the
HNE- (but not malondialdehyde-) modified apoB in LDL [171]. case, with relative levels of oxidized amino acids in proteins
Thus one might expect that oxidized proteins in general are resembling their extents of generation during HOd attack on
efficiently removed from cells and organisms, assuming that they proteins. Generally, hydroxylated aromatic residues are present
are accessible to appropriate proteinases. at higher levels than aliphatic residues. Dityrosine is present at
This argument has lead to several proposals that features of intermediate levels, and nitrotyrosine at low levels. This suggests
oxidized proteins constitute ‘ signals ’ or ‘ markers ’ for rapid that protein oxidation in ŠiŠo is a heterogeneous process giving
10 R. T. Dean and others

Light, irradiation, electron leaks, metals, inflammation, autoxidation

Radical generation

Antioxidant enzymes
Sequestration of transition metals Scavengers
Competing substrates Sacrificial antioxidants

PROTEIN OXIDATION

Protein radicals Aldehydes

Antioxidant removal Protein-bound reactive species

Chain

Reduction Derivatization
Secondary radicals

DEGRADATION STABLE END-PRODUCTS

OF PROTEIN OXIDATION

Marker
Excretion Oxidized amino
acids
Re-utilization (protein synthesis)

Figure 1 Postulated mechanisms of protein oxidation in vivo

Processes depicted in colour are those likely to result in the amelioration of protein damage in vivo.

rise to multiple products (as would be expected with HOd) rather value given the specific amino acid modifications shown in Table
than a highly selective one, and that all oxidized proteins are 1. The simplest explanation is that the carbonyl assay over-
removed in parallel, so that there is no selective removal, for estimates protein oxidation in complex biological materials ; this
example, of those protein molecules containing hydroxyvaline in assay may be best restricted to purified proteins.
comparison with those containing dopa. The hydroxy derivatives of leucine and valine can arise from
In aging and certain pathologies (discussed below), oxidized HOd attack. In contrast, 3-nitrotyrosine formation must involve
proteins accumulate, as has been demonstrated for inactive reactive nitrogen intermediates, but nitric oxide, peroxynitrite,
enzymes in nematodes [188]. This must indicate either physical nitrite and reactions between hypochlorite and nitrogen-con-
separation of the oxidized molecules from proteolytic systems (as taining compounds are just some possible sources of such
may apply in the lens), or that the rates of production and intermediates. Several of these species can also give rise to both
removal strike a new balance, while remaining mutually ac- hydroxylated aromatic residues and tyrosyl (phenoxyl) radicals,
cessible. Claims of altered activities of individual degradative and hence dityrosine. Dityrosine can also be formed by the
systems (e.g. the alkaline proteinases ; [189]), even when con- myeloperoxidase}chloride}H O system, with either protein-
# #
firmed, cannot be immediately interpreted as causal, since the bound or free tyrosine, as judged by model experiments. Were
routes of catabolism of oxidized proteins are not established, and this to contribute in ŠiŠo, then dityrosine formation might be
may well be pleiotropic. paralleled by 3-chlorotyrosine formation. Available data are
Fig. 1 summarizes the mechanisms by which oxidized proteins complicated by the possible further oxidation of dityrosine [192].
may undergo further reactions in ŠiŠo. It is also necessary to bear in mind the possibility of the
artefactual generation of oxidized proteins during processing
(e.g. [193]) ; this is a particular danger with post-mortem samples.
PHYSIOLOGY AND PATHOLOGY OF PROTEIN OXIDATION There are some physiological situations, such as the byssi of
It will ultimately be important to gain detailed information as to marine organisms from mussels to tunicates, in which oxidized
the pathways of protein oxidation in ŠiŠo ; this is difficult at proteins are clearly functionally important as adhesives. These
present. Protein carbonyl measurements [190] detect an in- unusual proteins contain very large portions of their tyrosine
completely characterized range of products, and are not straight- residues as dopa, as welll as further hydroxylated species. The
forward with tissue samples [189,191]. The commonly observed specificity of the (apparently non-enzymic) oxidations is lower
tissue levels of around 1 nmol}mg of protein suggest that 5 % of than that responsible for the formation of dopa or topa in active
proteins contain a carbonyl function, and roughly one carbonyl sites of certain redox enzymes. In accordance with this decreased
group per 3000 parent amino acids. This is a remarkably high specificity of oxidation, a variety of other unusual modified
 Radical-mediated protein oxidation 11

amino acids are also found in these marine proteins, and some ratios and excretion of GSSG [175,177,204–206]. Protein oxi-
function as efficient metal chelators (see [194–196]). dation (measured by carbonyl groups) is kinetically and spatially
correlated with these events. Reversible elevation of protein
carbonyl levels (from high basal values of around 10 nmol}mg of
Oxidized proteins in the control of cell growth and differentiation protein) by up to 2.5-fold in less than 1 h was observed in
Radicals are probably necessary mediators of cell growth control association with each transition [174]. Reversal also took less
[197] and heat-shock protein synthesis [198]. The question is than 1 h. The cell-free oxidative inactivation of glutamate de-
whether protein oxidation is an important factor in such hydrogenase and glutamine synthetase from the fungus [51,207]
mechanisms. Several transcription factors are redox-controlled, created more acidic polypeptides, and importantly such acidic
notably AP-1 and nuclear factor-κB. AP-1 may be important in species were also observed in ŠiŠo and found to display enhanced
regulation of cell growth [199], and the binding of AP-1 to DNA susceptibility to proteinases. Loss of glutamine synthetase activity
in 3B6 lymphocyte nuclear extracts is decreased when the cells occurred shortly before each differentiation step, coincident with
are growth-inhibited by selenite or selenodiglutathione, which the increase in protein carbonyls [176]. Total protein turnover
also oxidize thioredoxin. The functions of AP-1 and nuclear was also quantified, revealing highly selective degradation in the
factor-κB are distinct, and may vary from system to system [200]. regions of elevated carbonyls. It seems that protein oxidation can
Protein oxidation products could be effectors of cell growth, therefore be a central mediator of the degradative}synthetic
but present evidence suggests simply a bystander role. Thus the remodelling of protein profiles.
level of protein carbonyls in senescent IMR-90 human fibroblast A possible involvement of protein oxidation in programmed
cells does not vary significantly with population doublings, even cell death (apoptosis ; see [208]) remains to be established. Some
while growth control changes drastically [201]. Similarly, only oxidized components, notably LDL, can initiate apoptosis [209].
fibroblasts taken from humans over the age of 60, or from The protein p53, which seems to be involved in the induction of
Werner’s and Progeria cases, show carbonyl levels higher than apoptosis (e.g. [210]), may be regulated by alterations in protein
those in normal controls [202]. thiols. Its action can be triggered by radiation exposure [211],
There is evidence that protein oxidation plays a significant role and glutathione and other antioxidants can protect against this.
in cell differentiation. Several bacterial systems possess O -
#
dependent mechanisms for remodelling their enzyme profile.
Early studies by Turnbough and Switzer [37,38] examined the
Protein oxidation in host defence and tissue catabolism
idea of metal-ion-catalysed inactivation of enzymes as a trigger Physiological inflammatory reactions involve the recruitment of
for proteolytic removal. Two fascinating papers [52,203] concern leucocytes to sites of wounding or infection, and connective
protein catabolism in differentiating Klebsiella aerogenes and K. tissue catabolism and repair. Both topics are discussed together
pneumoniae respectively. The argument is put forward [203] that here, although it should be noted that, in chronic inflammation,
inactivation of aspartokinases III (lysine-sensitive) and I precede connective tissue breakdown can become pathological. Leucocyte
and lead to their enzymic proteolysis when K. aerogenes is recruitment usually occurs in response to chemotactic factors, a
incubated in a nitrogen-lacking medium. However, a detailed large number of which also stimulate radical production by the
analysis of these data and earlier immunological studies [39] leucocyte respiratory burst. These radicals play roles in cyto-
showed only that the two processes occur very close together. toxicity against infecting organisms, and perhaps also in host
Although lysate experiments indicate that both metal-ion- connective tissue catabolism.
catalysed inactivation and proteolysis can contribute to enzyme Ionic pump proteins involved in cellular homoeostasis are
inactivation, evidence that oxidative inactivation is responsible in often primary targets in the lysis of infecting organisms by
ŠiŠo is lacking. radicals produced by activated leucocytes. Subsequently,
These same issues were subsequently carefully addressed [52] neutrophil-generated HOCl can induce efficient target bacterial
with reference mainly to glycerol dehydrogenase, which is cell protein fragmentation [212]. Thomas and colleagues
inactivated and degraded when K. pneumoniae is moved from [213–215] found that free and protein thiol groups in the bacteria
anaerobic to aerobic conditions. The enzyme was purified from were very sensitive to hypochlorite, and that N-Cl compounds
cells before and after exposure to H O , which also inactivates it were long-lasting aggressive products, functioning even after the
# #
in intact cells. The purified inactivated enzyme showed enhanced supply of H O (and hence hypochlorite) had ceased. They also
# #
susceptibility to proteolysis by subtilisin, but there was little clear observed protein fragments, and up to 50 % of HOCl provided
evidence that it was oxidized. Chloramphenicol, an inhibitor of could be recovered as N-Cl-peptides.
protein synthesis (and usually of protein degradation), prevented In the degradation of proteins of the extracellular matrix, there
the inactivation and degradation of the glycerol dehydrogenase often seems to be an interaction between proteinase and radical
in cells exposed to O , but not in those exposed to H O , systems of the leucocytes, operative near to the cell surface
# # #
suggesting that some newly synthesized protein is necessary for [216,217]. For example, exposing glomerular basement
inactivation and}or degradation triggered by O . Whether this is membranes to myeloperoxidase}H O }chloride can enhance their
# # #
a protein component due to metal-catalysed oxidation, or a subsequent degradation. Model systems suggest that this is
proteinase, is not clear. Thus the question of protein oxidation as largely proteolytic, as antioxidants, scavengers and inhibitors of
a controlling step in the initiation of differentiative protein the respiratory burst have little effect. However, released oxidants
turnover remains to be settled. inactivate secreted elastase, lysozyme and β-glucuronidase, and
Conidiation in Neurospora crassa is another differentiation hence overall unchanged degradation was achieved with smaller
system in which the role of protein oxidation has been studied. amounts of enzyme, which could reflect co-operation with
Conidiation comprises successively the germination of the oxidants [217,218]. Others observed that PMA-triggered macro-
conidium (an asexual spore) to generate filamentous hyphae by phages degrade collagen suspensions by a brief radical burst
apical growth ; later, aerial hyphae develop, and ultimately followed by proteolysis ; the latter was restricted by proteinase
conidia form at their tips. There are thus three transitions, and inhibitors, the former by ascorbate or superoxide dismutase
each involves a ‘ hyperoxidant ’ state, as judged by chemilumi- [219].
nescence, alteration of GSH}GSSG and NAD(P)H}NAD(P)+ There are also examples of claimed oxidative activation of
12 R. T. Dean and others

proteolytic enzymes, such as urokinase by chloramines and aged tissues (see [237]). The key question is whether protein
hypochlorite [220]. There is good evidence for limited protein oxidation is primary or secondary in aging.
oxidation during inflammatory reactions ; for example, broncheo- Protein carbonyl levels increase in several gerbil tissues with
alveolar fluids have elevated methionine sulphoxide}methionine aging [234]. In contrast, increases in human fibroblasts aged in
ratios under conditions in which leucocyte activation is apparent Šitro or newly taken from older people are, respectively, nil [201]
[221]. Furthermore, oxidized proteins (and DNA) form inside or slight [202]. Glutamine synthetase activity declines and inactive
the leucocytes during their triggering, as judged by protein enzyme accumulates with aging [238]. However, there is little
carbonyls [222] and tyrosine modification [223], although it is not evidence that the inactive molecules are oxidized. The observation
clear whether this affects protein function directly. [239] that administration of the spin-trap PBN to rats prevents
the age-related accumulation of protein carbonyls, loss of glu-
Oxidized proteins and toxicity to the host organism tamine synthetase and neutral proteinase activities, and be-
havioural decrements is encouraging, but requires confirmation
Proteins are probably the most critical target of toxic damage, (see [189,240]). It seems unlikely that the levels of PBN achieved
since their inactivation can have rapid and supra-stoichiometric in this study would have any significant antioxidant effect.
effects, by virtue of their catalytic function. The inactivation of Long-lived proteins of specialized tissues such as the eye
a small portion of an ion transporter (in the plasma membrane permit the clearest indication that specific protein oxidation
or mitochondria) is likely to have much more drastic consequences products, such as dopa, dityrosine, o- and m-tyrosine, and valine
than a similarly extensive alteration of sterol molecules. and leucine hydroxides, do accumulate (S. Fu, R. Dean, M.
Most cytolysis ultimately depends on osmotic swelling and cell Southan and R. Truscott, unpublished work) in human catarac-
rupture, in contrast with the shrinkage mechanism of pro- tous lenses in association with loss of cysteine and methionine
grammed cell death. Lipid damage may be separable from (e.g. [241]). In human non-cataractous lenses there seems to be
critical events, as lipophilic antioxidants often restrict lipid little accumulation with age of o-tyrosine, but up to 33 %
oxidation without decreasing cell lysis. Protein damage, including increases in dityrosine, as well as vast increases in ‘ di-tyrosine ’
alterations of ion pumps, has been observed in a few cases. fluorescence [242].
Changes in membrane potential may be observed very rapidly In other tissues it is not clear how rates of production and
and, after a lag of 1–2 h, cell death ensues [224]. removal of oxidized protein relate to the elevated levels measured ;
The oxidation of ionic transporters has been closely studied in proteins in extracts of gerbil tissues show increased sensitivity to
relation to cardiac arrhythmias and ischaemia}reperfusion. An radiolytic oxidation with aging [243], but this is not the case for
ischaemic period is followed by rapid re-oxygenation [225] and rat tissue extracts undergoing endogenous or metal-ion-catalysed
radical fluxes, as well as by potentially damaging leucocyte oxidation [244]. Since the mechanisms of degradation of oxidized
infiltration. Elegant systems for studying oxidative affronts to proteins are not established, it is perhaps premature to relate
the heart and to cardiac cells have been developed, some using changes in the activity of individual proteolytic systems to the
photosensitizers to localize oxidation, and}or to permit "O accumulation of oxidized proteins in aging (cf. [245]).
#
production [226]. Thus, after perfusion of hearts with the Drosophila has long been used in aging studies, and shows
photosensitizer Rose Bengal, subsequent illumination leads, lipofuscin accumulation. Elegant studies by Sohal and co-
within seconds, to alterations in membrane transport and workers [246] showed an age-related increase in oxidant pro-
arrhythmias [227]. These events can be inhibited by antioxidants, duction in mitochondria of the flight muscle, together with
and membrane protein -SH oxidation occurs, although the increases in protein carbonyls and indices of DNA damage. Thus
precise lesion has not been defined. The data are consistent with a vicious circle, in which mitochondrial oxidative damage
primary damage to the Na+}H+ exchanger [228]. The importance enhances further oxidant production and damage, may operate.
of calcium transporters, notably the Na+}Ca#+ exchanger which Transgenic fly studies showed that overexpression of either
can be inactivated through its -SH group [229], is also apparent. catalase or superoxide dismutase singly had no effect on life
Alterations in cellular calcium may influence intracellular pro- span, whereas the simultaneous overexpression of both enzymes
teolysis and many other physiological systems. prolonged it [247,248]. This permitted an increased aggregate
Reperfusion of an ischaemic tissue can be accompanied by the physical activity, and was associated with lower levels of protein
release of redox-active transition-metal ions [230] and a burst of carbonyls at various ages. The flies consumed enhanced amounts
radicals [231] whose formation requires O . In working rat of O in the last two-thirds of their lives, and 30 % more overall.
# #
hearts, reperfusion is accompanied by a large rise in protein The overexpression of antioxidant enzymes is expected to help
carbonyls within 5 min, followed by a return within 15 min the flies resist the increased oxidative flux, but the changed O
almost to basal levels [232]. Similar observations have been made #
consumption with age suggests that there are other factors
in relation to brain ischaemia and reperfusion ; some antioxidants involved. Whether the decreased pool sizes of oxidized proteins
[233] and the spin trap N-t-butyl-α-phenylnitrone (PBN) [234] reflect altered generation or removal has not been directly tested.
may decrease this. Several other antioxidant regimes can restrict The levels of lipid oxidation products do not rise continuously
overall damage. However, the precise interaction of these restrict- with age [249], suggesting that either oxidative events or product
ing regimes with protein oxidation is unknown, and there are removal must not be linear with time. It remains to be established
disparities (see [235]) that require clarification. Site-specific whether protein oxidation is critical in the progression of aging,
modifications of selected proteins have been observed [236], and but these studies encourage further investigation (see [250]).
could explain these disparities.

Diabetes
PROTEIN OXIDATION IN AGING AND IN SELECTED PATHOLOGIES
Diabetes is one of the most common inherited diseases, caused
Aging by the impaired production of insulin by pancreatic islet β-cells
Inactive forms of several proteins accumulate in aging cells, as and}or by diminished tissue responses to insulin (insulin re-
shown for the nematode Turbatrix aceti [188]. ‘ Age pigments ’ sistance). The consquence of this is that circulating blood glucose
that contain protein, such as lipofuscin, also accumulate in some levels are chronically elevated, and this appears to be mainly
 Radical-mediated protein oxidation 13

Table 2 Selected data on concentrations of oxidized amino acids in tissue proteins from humans under normal and pathological circumstances
‘Unpublished work ’ refers to work of the present authors ; results are means from three advanced human atherosclerotic plaques and three human LDL samples obtained fresh. In these studies,
levels of the measured species in fresh human plasma expressed per parent amino acid were in most cases very slightly lower. The aorta samples used in the studies of Heinecke et al. [310]
were obtained post mortem. ND, detailed data not found.

Product Physiological levels Pathological levels Source

Dopa 85 pmol/mg of LDL protein 410 pmol/mg of protein (14/10000 tyrosines) Unpublished work
(6/10000 tyrosines) in advanced human atherosclerotic plaques
o- and m-tyrosine 62 and 35 pmol/mg of LDL protein 105 and 175 pmol/mg of protein Unpublished work
(5 and 3/10000 phenylalanines) (3.5 and 6/10000 phenylalanines)
respectively respectively in plaques
No increase in atherosclerotic aorta samples [310]
compared with normal
5/10000 phenylalanines Unchanged with age [242]
in human lenses of any age
N-Formylkynurenine ; kynurenine ND ND
Dityrosine 0.2 pmol/mg of LDL protein 150 pmol/mg of protein (5/10000 tyrosines) Unpublished work
(0.02/10000 tyrosines) in plaques
10-fold elevated in aortic lesions in comparison [310]
with normal aortic samples ; 0.03/10000
tyrosines in lens proteins from old people
0.01/10000 tyrosines in human [242]
lens proteins from young people
Dimers of hydroxylated ND ND
aromatic amino acids
2-Oxohistidine ND ND
Hydro(pero)xyleucine β-Hydroxyleucine 2 : 5 pmol/mg of β-Hydroxyleucine 2 : 20 pmol/mg of Unpublished work
LDL protein (0.1/10000 leucines) protein (0.2/10000 leucines)
Hydro(pero)xyvalines β-Hydroxyvaline 1 : 5 pmol/mg of β-Hydroxyvaline 1 : 10 pmol/mg of protein Unpublished work
LDL protein (0.1/10000 valines) (0.1/10000 valines) in plaques
3-Chlorotyrosine Normal aorta : 0.8/10000 tyrosines Atherosclerotic aorta : 4.2/10000 tyrosines [310]
3-Nitrotyrosine ! 10 pmol/mg of LDL protein ! 10 pmol/mg of protein Unpublished
work*
(! 1/10000 tyrosines) (! 0.3/10000 tyrosines) in plaques
100-fold elevated in aortic lesion LDL [310]
compared with normal plasma LDL
p-Hydroxyphenylacetaldehyde ND ND
Aminomalonic acid 0.04–0.3/1000 total amino acids 0.2/1000 glycines in post-mortem [311]
in two E. coli strains human plaque
5-Hydroxy-2-aminovaleric acid 0.15 nmol/mg of protein in 100 000 g Unchanged in old mouse livers ; [312]
supernatants of young mouse livers elevated by hyperoxic exposure
Protein carbonyls C 1 nmol/mg of protein in many % 8 nmol/mg of protein in [190,191]
physiological tissue samples diseased brain samples
* Below UV-HPLC detection limit.

responsible for the major problems of diabetes, i.e. blindness, [27,28,255]. We have also demonstrated substantial formation of
gangrene and kidney failure. a range of aliphatic amino acid hydroxides, dopa, dityrosine and
Glycation reactions between glucose and proteins and other hydroxylation products of phenylalanine during exposure of
biological molecules, notably the Maillard (browning) reaction, proteins to glucose (S. Fu, M.-X. Fu, J. W. Baynes, S. R. Thorpe
have long been considered a potential major part of these and R. T. Dean, unpublished work).
consequences. ‘ Autoxidative glycosylation ’, involving autoxida- Glyoxal and arabinose are major autoxidation products of
tion of glucose via dicarbonyl intermediates and of protein- glucose ; the former can form N-ε-(carboxymethyl)lysine [256]
bound sugars, was proposed to be important [251,252]. Here we and the latter can participate in the generation of the fluorescent
only consider the specific relevance of protein oxidation. pentosidine cross-links in proteins [257]. Ascorbate, and
Metal-ion-catalysed autoxidation of sugars and products of dicarbonyl sugars such as methylglyoxal and 3-deoxyglucosone,
protein glycation can generate radicals that initiate and propagate which are observed in ŠiŠo, may participate in autoxidative
protein damage [128,251–253]. Cross-linking of collagen in Šitro reactions contributing to browning. There are also other sources
by exposure to glucose is virtually blocked under N and in the of glyoxal and related aldehydes such as those from amino acids
#
presence of chelators, confirming the importance of oxidative (e.g. [1]). The protein-bound sugar Amadori intermediates are
events [254]. We have shown that proteins exposed to generally more readily autoxidized than free sugars, so the
autoxidizing glucose contain protein-bound oxidizing and re- relative importance of autoxidation before and after protein
ducing species, later characterized as hydroperoxides and dopa binding may vary with the relative concentrations of the com-
14 R. T. Dean and others

ponents. N-ε-(Carboxymethyl)lysine and pentosidine accumulate this regard, and carbonyl levels correlate well with tangles [273].
in ŠiŠo with aging, and their levels can be elevated in diabetes and Glycation and autoxidative glycosylation products also accumu-
restricted in animal models by certain antioxidant regimes (see late in Alzheimer’s brains [274]. Importantly, the accumulation
[70]). Current evidence points towards a significant role for sugar of oxidation products in particular regions of the brain seems to
autoxidation, be it of free or protein-bound sugars, in the be related to specific cognitive defects [275]. In line with this,
complications of diabetes. administration of the spin-trap PBN to gerbils retards both
protein oxidation and such neurological defects (e.g. [239,276]).
How might these oxidized proteins act ? Amyloid β-protein is
Atherosclerosis cytotoxic, and induces oxidative damage in some brain cells
Diabetes predisposes to the age-related vascular disease athero- [277]. This may involve a receptor response mechanism, although
sclerosis, and this may be partly because of oxidative features. it has been claimed that the protein itself can generate radicals
Atherosclerosis involves vessel wall deposition of cells and lipids, [278], even though freshly dissolved samples were inactive [279].
forming the hallmark ‘ foam cells ’, and leads to cellular pro- Variations between samples (e.g. [278]) and upon ‘ aging ’ of the
liferation, wall thickening and calcium deposition. It is well materials make one wary of these data, as do the detailed
established that macromolecule oxidation accompanies athero- characteristics of the EPR spectra. Regardless of this problem,
sclerosis ; for example, around 30 % of the cholesterol linoleate radical metabolism [280] and protein oxidation could lead either
accumulated in atherosclerotic vessel walls (it is lacking from to cross-linking and aggregation of proteins into tangles or to
normal walls) is oxidized [258]. Much of the accumulated lipid cytotoxic signals that alter ion transporters and neuro-
seems to come from LDL, and the modification of its protein transmission [281], and in these ways may contribute specifically
apoB by aldehydes and lipid oxidation has been argued to be to the progression of Alzheimer’s disease.
central in lipid deposition in atherosclerosis [259]. It is necessary
to distinguish between direct protein oxidation and derivatiza-
tions secondary to lipid oxidation.
OUTLOOK : ON THE CONTROL OF PROTEIN OXIDATION
Here we consider only the evidence concerning direct apoB Mechanisms for the control of protein oxidation have been
oxidation. ApoB in LDL can be readily oxidized and fragmented rather little studied. No substantial attempt has yet been made to
by radiolysis [123] or by metal-ion-catalysed oxidation [149]. In elucidate the features of protein antioxidation that might be
ŠiŠo evidence concerning oxidized apoB is limited, given distinct from those of lipid antioxidation, e.g. as a result of the
ambiguities concerning 3-nitrotyrosine disposition between apoB relatively different importance of peroxyl and alkoxyl radicals in
and other proteins in plaque [260] and the low levels of this the two processes. Factors such as control of radical generation,
moiety and other oxidized amino acids in proteins (see Table 2), disposition of targets in relation to radical fluxes and breaking
and given the focus on derivatized rather than directly oxidized chain reactions seem likely to be important for both cases. Many
proteins (e.g. [261]). Studies on hypochlorite-oxidized protein, in of these basic mechanisms are discussed in earlier reviews [7,265].
which antisera specificities have been carefully validated, suggest Restricting the availability of transition-metal ions for oxi-
not only that hypochlorite-oxidized proteins are present in dative interconversions is one key factor in antioxidation. In this
plaques, but also that apoB peptides are among them [262]. respect proteins are often ambivalent, since they often present
HOCl appears to attack the protein selectively [97,263]. suitable binding sites, in the form of combinations of histidine,
If oxidized proteins are present in plaques, then we need to cysteine and other residues, but by binding metal ions they may
assess their relative rates of production and removal, and not either restrict the redox activity of the metal [126,282] or, if not,
simply to rely on pool sizes. Oxidative mechanisms may include localize damage upon themselves [283].
metal-ion-dependent cellular processes, or putatively metal- A candidate specialized protein for the protection of protein
independent ones, such as the action of HOCl. Yet extracellular thiol groups has been identified [284,285]. It restricts metal-ion-
fluids contain antioxidant and chelator capacity [264], and dependent protein oxidation involving thiols, but not systems in
advanced plaques themselves are not deficient in aqueous or which thiol is replaced by ascorbate. The protein may be an
lipophilic antioxidants such as ascorbate and tocopherol [258]. analogue of superoxide dismutase, which dismutes RSSR−d
Tocopherol-mediated peroxidation may be involved (see radicals, although it may also control metal-ion reduction (but
[265,266]). The inter- and}or intra-cellular site(s) of oxidation not simply sequester metals). This protein may collaborate with
are not clear, and may require specialized conditions. A myriad the reversible thiolation mechanism discussed above.
of potentially pathogenic activities have been attributed to LDL Enzymic antioxidants, such as superoxide dismutase and
and to some of its components at various poorly defined stages catalase, and glutathione and phospholipid peroxidase, can
of oxidation (e.g. [267]). The possible roles of the oxidized remove peroxides and intermediate radicals. Sacrificial anti-
protein components need to be considered. oxidants, such as ascorbate, urate, vitamin E and quinols, can
limit peroxidation chains by removing intermediate radicals,
such as peroxyl radicals, and themselves become radical species,
Neurodegenerative diseases which can eventually be detoxified. For example, Trolox C can
Toxic roles for oxidized proteins, in preference to oxidized lipids, restrict radical damage to albumin [147]. There are indications of
have been proposed in recent literature on Alzheimer’s disease specialized protective mechanisms for proteins, notably bilirubin
[268], even though the latter also accumulate [269]. In particular, present on albumin, that may decrease protein damage by
the accumulation of aggregated amyloid β-protein in diseased repairing protein radicals or by scavenging externally presented
brains as neurofibrillary tangles can occur through oxidation radicals [22]. When proteins are present in membranes or
[270], and mitochondrial dysfunction has been envisaged as a lipoproteins, agents (such as vitamin E plus vitamin C) that
radical source [271]. restrict lipid oxidation may consequently restrict protein oxi-
The content of protein carbonyls in Alzheimer’s brain samples dation, but it is unclear whether any also directly interact with
is greater than in age-matched controls [272], and this provides protein radicals (cf. discussion of ubiquinol [286]).
the clearest indication of greater accumulation of oxidized The reactive intermediates (besides protein radicals) in protein
proteins in this disease. Brain regions show specific changes in oxidation, such as the protein hydroperoxides, may also be
 Radical-mediated protein oxidation 15

selectively detoxified, and repair of some subsequent inter- 11 Wagner, A. F., Frey, M., Neugebauer, F. A., Schafer, W. and Knappe, J. (1992) Proc.
mediates has been discussed above. For example, protein hydro- Natl. Acad. Sci. U.S.A. 89, 996–1000
peroxides are reduced efficiently by cells to stable (non-radical- 12 Kopoldova, J., Liebster, J. and Babicky, A. (1963) Int. J. Appl. Radiat. Isot. 14,
455–460
generating) hydroxides, and there seems to be a cell surface 13 Fu, S., Hick, L. A., Sheil, M. M. and Dean, R. T. (1995) Free Radical Biol. Med. 19,
mechanism for this ; one can presume that there are intracellular 281–292
counterparts also [29]. Metal-ion-mediated reduction can, how- 14 Dean, R. T., Fu, S., Gieseg, G. and Armstrong, S. G. (1996) in Free Radicals : A
ever, also be accompanied by further radical fluxes [25,287], so at Practical Approach (Punchard, N. A. and Kelly, F. J., eds.), pp. 171–183, IRL Press,
present it is not clear how protection (direct reduction) and Oxford
radical generation are balanced. Lipoprotein lipid hydro- 15 Aeschbach, R., Amado, R. and Neukom, H. (1976) Biochim. Biophys. Acta 439,
292–301
peroxides are similarly reduced both in lipoproteins [288] and by
16 Amato, R., Aeschbach, R. and Neukom, H. (1984) Methods Enzymol. 107, 377–388
cells [289]. 17 Dale, W. M., Davies, J. V. and Gilbert, C. W. (1949) Biochem. J. 45, 93–99
Thioredoxin plus thioredoxin reductase can reduce protein 18 Jayson, G. G., Scholes, G. and Weiss, J. (1954) Biochem. J. 57, 386–390
disulphides to thiols, including those formed by the thiol-specific 19 Uchida, K. and Kawakishi, S. (1993) FEBS Lett. 332, 208–210
antioxidant enzyme upon reduction of alkyl hydroperoxides 20 Raper, H. S. (1932) Biochem. J. 26, 2000–2004
[290]. This activity is thought to be responsible for the in- 21 Neuzil, J., Gebicki, J. M. and Stocker, R. (1993) Biochem. J. 293, 601–606
volvement of the enzyme in the regulation of redox-sensitive 22 Neuzil, J. and Stocker, R. (1993) FEBS Lett. 331, 281–284
23 Barron, E. S. G., Ambrose, J. and Johnson, P. (1955) Radiat. Res. 2, 145–158
transcription factors. In addition, thioredoxin reductase and 24 Davies, K. J., Delsignore, M. E. and Lin, S. W. (1987) J. Biol. Chem. 262,
thioredoxin have alkyl hydroperoxide reductase activity and are 9902–9907
efficient electron donors to plasma glutathione peroxidase [291]. 25 Davies, M. J., Fu, S. and Dean, R. T. (1995) Biochem. J. 305, 643–649
It is not known whether this reductase acts on protein hydro- 26 Davies, M. J. (1996) Arch. Biochem. Biophys. 336, 163–172
peroxides, although glutathione peroxidase and human plasma 27 Simpson, J. A., Narita, S., Gieseg, S., Gebicki, S., Gebicki, J. M. and Dean, R. T.
components can do so [29]. As discussed above, thiol-specific (1992) Biochem. J. 282, 621–624
28 Gieseg, S. P., Simpson, J. A., Charlton, T. S., Duncan, M. W. and Dean, R. T. (1993)
antioxidant enzyme also protects proteins against thiyl-radical-
Biochemistry 32, 4780–4786
induced inactivation [292], although it is not known whether 29 Fu, S., Gebicki, S., Jessup, W., Gebicki, J. M. and Dean, R. T. (1995) Biochem. J.
thiyl radicals on proteins are substrates for it. 311, 821–827
In looking to future research and preventative prospects, 30 Latarjet, R. and Loiseleur, J. (1942) C. R. Seances Soc. Biol. Fil. 136, 60–63
several important aspects need to be considered regarding non- 31 Gebicki, S. and Gebicki, J. M. (1993) Biochem. J. 289, 743–749
proteinaceous antioxidants for proteins. First, with the exception 32 Aldrich, J. E., Cundall, R. B., Adams, G. E. and Willson, R. L. (1969) Nature (London)
of cysteines, protein amino acids are less reactive than lipids 221, 1049–1050
33 Adams, G. E., Bisby, R. H., Cundall, R. B., Redpath, J. L. and Willson, R. L. (1972)
containing bisallylic hydrogens, so that reactive radical oxidants Radiat. Res. 49, 290–299
(HOd, ROd) are of greater relevance to protein than to lipid 34 Adams, G. E., Posener, M. L., Bisby, R. H., Cundall, R. B. and Key, J. R. (1979) Int.
(per)oxidation. Since the differences in rate constants of a radical J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 35, 497–507
with its targets decrease with increasing reactivity of the radical, 35 Terato, H. and Yamamoto, O. (1994) Biochem. Mol. Biol. Int. 34, 295–300
it follows that the conventional scavenging antioxidants (α- 36 Anderson, R. F., Patel, K. B. and Adams, G. E. (1977) Int. J. Radiat. Biol. Relat. Stud.
tocopherol, ascorbate, urate, etc.) are less likely to offer specific Phys. Chem. Med. 32, 523–531
37 Turnbough, C. L. and Switzer, R. L. (1975) J. Bacteriol. 121, 108–114
protection to proteins than they do to lipids.
38 Turnbough, C. L. and Switzer, R. L. (1975) J. Bacteriol. 121, 115–120
Secondly, the chain reaction of lipid peroxidation is probably 39 Levine, R. L., Oliver, C. N., Fulks, R. M. and Stadtman, E. R. (1981) Proc. Natl. Acad.
the result of the predominant presence of a single chain-carrying Sci. U.S.A. 78, 2120–2124
species ; in lipoproteins, α-tocopheroxyl radical is the likely chain 40 Nakamura, K. and Stadtman, E. R. (1984) Proc. Natl. Acad. Sci. U.S.A. 81,
carrier as long as vitamin E is present [293], whereas, after 2011–2015
tocopherol depletion, lipid peroxyl radicals become the chain- 41 Stadtman, E. R. (1991) Am. J. Clin. Nutr. 54, 1125S–1128S
carrying species. Thus an antioxidant for lipids can focus on the 42 Fucci, L., Oliver, C. N., Coon, M. J. and Stadtman, E. R. (1983) Proc. Natl. Acad. Sci.
U.S.A. 80, 1521–1525
elimination of a single reactant, independent of the nature of the 43 Stadtman, E. R. and Chock, P. B. (1978) Curr. Top. Cell. Regul. 13, 53–95
peroxidation-initiating radical [293]. In contrast, there are several 44 Stadtman, E. R. (1990) Methods Enzymol. 182, 793–809
chain-carrying species invloved during radical-induced protein 45 Rivett, A. J., Roseman, J. E., Oliver, C. N., Levine, R. L. and Stadtman, E. R. (1985)
(per)oxidation, such that individual antioxidants may perhaps be Prog. Clin. Biol. Res. 180, 317–328
expected only to attenuate the process. On the other hand, the 46 Kim, K., Rhee, S. G. and Stadtman, E. R. (1985) J. Biol. Chem. 260, 15394–15397
apparently more substantial role of alkoxyl radicals in protein 47 Cervera, J. and Levine, R. L. (1988) FASEB J. 2, 2591–2595
48 Levine, R. L. (1983) J. Biol. Chem. 258, 11823–11827
than in lipid peroxidation may offer at least a selective target for
49 Rivett, A. J. and Levine, R. L. (1990) Arch. Biochem. Biophys. 278, 26–34
inhibition of protein rather than lipid oxidation. 50 Climent, I. and Levine, R. L. (1991) Arch. Biochem. Biophys. 289, 371–375
51 Aguirre, J. and Hansberg, W. (1986) J. Bacteriol. 166, 1040–1045
52 Chevalier, M., Lin, E. C. and Levine, R. L. (1990) J. Biol. Chem. 265, 40–46
REFERENCES 53 Thomas, R. M., Nauseef, W. M., Iyer, S. S., Peterson, M. W., Stone, P. J. and Clark,
1 Dakin, H. D. (1906) J. Biol. Chem. 1, 171–176 R. A. (1991) J. Leukocyte Biol. 50, 568–579
2 Dakin, H. D. (1908) J. Biol. Chem. 4, 63–76 54 Estell, D. A., Graycar, T. P. and Wells, J. A. (1985) J. Biol. Chem. 260, 6518–6521
3 Hopkins, F. G. (1925) Biochem. J. 19, 787–819 55 Kwon, N. S., Chan, P. C. and Kesner, L. (1990) Agents Actions 29, 388–393
4 Bevilotti, V. (1945) Boll. Soc. Ital. Biol. Sper. 20, 128–129 56 Dean, R. T., Nick, H. P. and Schnebli, H. P. (1989) Biochem. Biophys. Res. Commun.
5 Drake, M. D., Giffee, J. W., Johnson, D. W. and Koenig, V. L. (1957) J. Am. Chem. 159, 821–827
Soc. 79, 1395–1401 57 Gantchev, T. G. and van Lier, J. (1995) Photochem. Photobiol. 62, 123–134
6 von Sonntag, C. (1987) The Chemical Basis of Radiation Biology, Taylor and Francis, 58 Gopalakrishna, R., Chen, Z. H. and Gundimeda, U. (1992) FEBS Lett. 314, 149–154
London 59 Pedersen, J. Z. and Finazzi, A. A. (1993) FEBS Lett. 325, 53–58
7 Halliwell, B. and Gutteridge, J. M. C. (1989) Free Radicals in Biology and Medicine, 60 Licht, S., Gerfen, G. J. and Stubbe, J. (1996) Science 271, 477–481
Clarendon, Oxford 61 Mao, S. S., Holler, T. P., Bollinger, J. J., Yu, G. X., Johnston, M. I. and Stubbe, J.
8 Rodgers, M. A., Sokol, H. A. and Garrison, W. M. (1968) J. Am. Chem. Soc. 90, (1992) Biochemistry 31, 9744–9751
795–796 62 Brush, E. J., Lipsett, K. A. and Kozarich, J. W. (1988) Biochemistry 27, 2217–2222
9 Garrison, W. M. (1968) Curr. Top. Radiat. Res. 4, 43–94 63 Cucurou, C., Battioni, J. P., Thang, D. C., Nam, N. H. and Mansuy, D. (1991)
10 Garrison, W. M. (1987) Chem. Rev. 87, 381–398 Biochemistry 30, 8964–8970
16 R. T. Dean and others

64 Mohsenin, V. and Gee, J. L. (1989) J. Appl. Physiol. 66, 2211–2215 112 Kang, J. O., Chan, P. O. and Kesner, L. (1985) Inorg. Chim. Acta 107, 253–258
65 Andersen, H. J., Chen, H., Pellett, L. J. and Tappel, A. L. (1993) Free Radical Biol. 113 Schuessler, H. and Herget, A. (1980) Int. J. Radiat. Biol. Relat. Stud. Phys. Chem.
Med. 15, 37–48 Med. 37, 71–80
66 Andersen, H. J., Pellett, L. and Tappel, A. L. (1994) Chem.-Biol. Interact. 93, 114 Schuessler, H. and Schilling, K. (1984) Int. J. Radiat. Biol. Relat. Stud. Phys. Chem.
155–169 Med. 45, 267–281
67 Chio, K. S. and Tappel, A. L. (1969) Biochemistry 8, 2827–2832 115 Puchala, M. and Schuessler, H. (1995) Int. J. Pept. Protein Res. 46, 326–332
68 Szweda, L. I., Uchida, K., Tsai, L. and Stadtman, E. R. (1993) J. Biol. Chem. 268, 116 Schuessler, H., Niemczyk, P., Eichhorn, M. and Pauly, H. (1975) Int. J. Radiat. Biol.
3342–3347 Relat. Stud. Phys. Chem. Med. 28, 401–408
69 Kautiainen, A. (1992) Chem.–Biol. Interact. 83, 55–63 117 Davies, K. J., Lin, S. W. and Pacifici, R. E. (1987) J. Biol. Chem. 262, 9914–9920
70 Baynes, J. W. (1996) in Drugs, Diet and Disease, vol. 2 (Ioannides, C., ed.), pp. 118 Dean, R. T., Wolff, S. P. and McElligott, M. A. (1989) Free Radical Res. Commun.
201–240, Pergamon, London 7, 97–103
71 Jessup, W., Jurgens, G., Lang, J., Esterbauer, H. and Dean, R. T. (1986) Biochem. 119 Uchida, K., Kato, Y. and Kawakishi, S. (1990) Biochem. Biophys. Res. Commun.
J. 234, 245–248 169, 265–271
72 Hoff, H. F., O’Neil, J., Chisolm, G. D., Cole, T. B., Quehenberger, O., Esterbauer, H. 120 Easton, C. J. (1991) in Advances in Detailed Reaction Mechanisms (Coxon, J. M.,
and Jurgens, G. (1989) Arteriosclerosis 9, 538–549 ed.), pp. 83–126, JAI Press, Greenwich, CT
73 Bruenner, B. A., Jones, A. D. and German, J. B. (1995) Chem. Res. Toxicol. 8, 121 Burgess, V. A., Easton, C. J. and Hay, M. P. (1989) J. Am. Chem. Soc. 111,
552–559 1047–1052
74 Nadkarni, D. V. and Sayre, L. M. (1995) Chem. Res. Toxicol. 8, 284–291 122 Monboisse, J. C., Braquet, P., Randoux, A. and Borel, J. P. (1983) Biochem. Pharm.
75 Uchida, K. and Stadtman, E. R. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 4544–4548 32, 53–58
76 Cohn, J. A., Tsai, L., Friguet, B. and Szweda, L. I. (1996) Arch. Biochem. Biophys. 123 Bedwell, S., Dean, R. T. and Jessup, W. (1989) Biochem. J. 262, 707–712
328, 158–164 124 Sperling, J. and Elad, D. (1971) J. Am. Chem. Soc. 93, 967–971
77 Uchida, K. and Stadtman, E. R. (1993) J. Biol. Chem. 268, 6388–6393 125 Sperling, J. and Elad, D. (1971) J. Am. Chem. Soc. 93, 3839–3840
78 McNamara, M. and Augusteyn, R. C. (1984) Exp. Eye Res. 38, 45–56 126 Marx, G. and Chevion, M. (1986) Biochem. J. 236, 397–400
79 Prutz, W. A. (1992) Int. J. Radiat. Biol. 61, 593–602 127 Hunt, J. V., Simpson, J. A. and Dean, R. T. (1988) Biochem. J. 250, 87–93
80 Cappiello, M., Voltarelli, M., Giannessi, M., Cecconi, I., Camici, G., Manao, G., Del, 128 Hunt, J. V., Dean, R. T. and Wolff, S. P. (1988) Biochem. J. 256, 205–212
Corso A. and Mura, U. (1994) Exp. Eye Res. 58, 491–501 129 Levitzki, A., Anbar, M. and Berger, A. (1967) Biochemistry 6, 3757–3765
81 Rokutan, K., Thomas, J. A. and Sies, H. (1989) Eur. J. Biochem. 179, 233–239 130 Orr, C. W. (1967) Biochemistry 6, 3000–3006
82 Seres, T., Ravichandran, V., Moriguchi, T., Rokutan, K., Thomas, J. A. and Johnston, 131 Orr, C. W. (1967) Biochemistry 6, 2995–3000
R. J. (1996) J. Immunol. 156, 1973–1980 132 Guptasarma, P., Balasubramanian, D., Matsugo, S. and Saito, I. (1992) Biochemistry
83 Smith, C. E., Stack, M. S. and Johnson, D. A. (1987) Arch. Biochem. Biophys. 253, 31, 4296–4303
146–155 133 Eiserich, J. P., Cross, C. E., Jones, A. D., Halliwell, B. and van der Vliet, A. (1996)
84 Boudier, C. and Bieth, J. G. (1994) Biochem. J. 303, 61–68 J. Biol. Chem. 271, 19199–19208
85 Keck, R. G. (1996) Anal. Biochem. 236, 56–62 134 Heinecke, J. W., Li, W., Daehnke, H. D. and Goldstein, J. A. (1993) J. Biol. Chem.
86 Banerjee, S. K. and Mudd, J. B. (1992) Arch. Biochem. Biophys. 295, 84–89 268, 4069–4077
87 Knight, K. L. and Mudd, J. B. (1984) Arch. Biochem. Biophys. 229, 259–269 135 Giulivi, C. and Davies, K. J. (1993) J. Biol. Chem. 268, 8752–8759
88 Berlett, B. S., Levine, R. L. and Stadtman, E. R. (1996) J. Biol. Chem. 271, 136 Marx, G. and Chevion, M. (1985) Thromb. Res. 40, 11–18
4177–4182 137 Shacter, E., Williams, J. A. and Levine, R. L. (1995) Free Radical Biol. Med. 18,
89 Rossi, F., De, T. P., Bellavite, P., Della, B. V. and Grzeskowiak, M. (1983) Biochim. 815–821
Biophys. Acta 758, 168–175 138 Rieser, P. (1956) Proc. Soc. Exp. Biol. Med. 91, 654–657
90 Bisby, R. H., Ahmed, S. and Cundall, R. B. (1984) Biochem. Biophys. Res. 139 Schacter, E., Williams, J. A., Lim, M. and Levine, R. L. (1994) Free Radical Biol.
Commun. 119, 245–251 Med. 17, 429–437
91 Brot, N., Weissbach, L., Werth, J. and Weissbach, H. (1981) Proc. Natl. Acad. Sci. 140 Gray, H. B. and Winkler, J. R. (1996) Annu. Rev. Biochem. 65, 537–561
U.S.A. 78, 2155–2158 141 Roubal, W. T. and Tappel, A. L. (1966) Arch. Biochem. Biophys. 113, 5–8
92 Moskovitz, J., Jenkins, N. A., Gilbert, D. J., Copeland, N. G., Jursky, F., Weissbach, 142 Tappel, A. L. (1973) Proc. Natl. Acad. Sci. U.S.A. 32, 1870–1874
H. and Brot, N. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 3205–3208 143 Gardner, H. W. (1979) J. Agric. Food Chem. 27, 220–229
93 Nielsen, H. K., Loliger, J. and Hurrell, R. F. (1985) Br. J. Nutr. 53, 61–73 144 Dean, R. T., Thomas, S. M. and Garner, A. (1986) Biochem. J. 240, 489–494
94 Vogt, W. (1995) Free Radical Biol. Med. 18, 93–105 145 Dean, R. T. and Cheeseman, K. H. (1987) Biochem. Biophys. Res. Commun. 148,
95 Yim, H. S., Kang, S. O., Hah, Y. C., Chock, P. B. and Yim, M. B. (1995) J. Biol. 1277–1282
Chem. 270, 28228–28233 146 Thomas, S. M., Gebicki, J. M. and Dean, R. T. (1989) Biochim. Biophys. Acta 1002,
96 Wells-Knecht, M. C., Thorpe, S. R. and Baynes, J. W. (1995) Biochemistry 34, 189–197
15134–15141 147 Dean, R. T., Hunt, J. V., Grant, A. J., Yamamoto, Y. and Niki, E. (1991) Free Radical
97 Hazell, L. J., van den Berg, J. J. and Stocker, R. (1994) Biochem. J. 302, 421–428 Biol. Med. 11, 161–168
98 Fisher, M. T. and Stadtman, E. R. (1992) J. Biol. Chem. 267, 1872–1880 148 Hunt, J. V. and Dean, R. T. (1989) Biochem. Biophys. Res. Commun. 162,
99 Wolff, S. P. and Dean, R. T. (1986) Biochem. J. 234, 399–403 1076–1084
100 Davies, K. J. (1987) J. Biol. Chem. 262, 9895–9901 149 Fong, L. G., Parthasarathy, S., Witztum, J. L. and Steinberg, D. (1987) J. Lipid Res.
101 Dean, R. T., Thomas, S. M., Vince, G. and Wolff, S. P. (1986) Biomed. Biochim. 28, 1466–1477
Acta 45, 1563–1573 150 Heinecke, J. W., Li, W., Francis, G. A. and Goldstein, J. A. (1993) J. Clin. Invest.
102 Wang, K. and Spector, A. (1995) Invest. Ophthalmol. Vis. Sci. 36, 311–321 91, 2866–2872
103 Prinsze, C., Dubbelman, T. M. and Van, S. J. (1990) Biochim. Biophys. Acta 1038, 151 Samuni, A., Chevion, M., Halpern, Y. S., Ilan, Y. A. and Czapski, G. (1978) Radiat.
152–157 Res. 75, 489–496
104 Walker, K. W., Lyles, M. M. and Gilbert, H. F. (1996) Biochemistry 35, 1972–1980 152 Hartman, P. S., Eisenstark, A. and Pauw, P. G. (1979) Proc. Natl. Acad. Sci. U.S.A.
105 Yamashita, Y., Shimokata, K., Mizuno, S., Daikoku, T., Tsurumi, T. and Nishiyama, 76, 3228–3232
Y. (1996) J. Virol. 70, 2237–2246 153 Albrich, J. M., McCarthy, C. A. and Hurst, J. K. (1981) Proc. Natl. Acad. Sci. U.S.A.
106 Teshima, S., Rokutan, K., Takahashi, M., Nikawa, T. and Kishi, K. (1996) Biochem. 78, 210–214
J. 315, 497–504 154 Quintanilha, A. T. and Davies, K. J. (1982) FEBS Lett. 139, 241–244
107 Ericsson, M., Tarnvik, A., Kuoppa, K., Sandstrom, G. and Sjostedt, A. (1994) Infect. 155 Parinandi, N. L., Zwizinski, C. W. and Schmid, H. H. (1991) Arch. Biochem.
Immun. 62, 178–183 Biophys. 289, 118–123
108 Dreher, D., Vargas, J. R., Hochstrasser, D. F. and Junod, A. F. (1995) 156 Reinheckel, T., Wiswedel, I., Noack, H. and Augustin, W. (1995) Biochim. Biophys.
Electrophoresis 16, 1205–1214 Acta 1239, 45–50
109 Puchala, M. and Schuessler, H. (1993) Int. J. Radiat. Biol. 64, 149–156 157 Richter, C., Park, J. W. and Ames, B. N. (1988) Proc. Natl. Acad. Sci. U.S.A. 85,
110 Davies, K. J. and Delsignore, M. E. (1987) J. Biol. Chem. 262, 9908–9913 6465–6467
111 Dean, R. T., Roberts, C. R. and Jessup, W. (1985) Prog. Clin. Biol. Res. 180, 158 Richter, C., Gogvadze, V., Laffranchi, R., Schlapbach, R., Schweizer, M., Suter, M.,
341–350 Walter, P. and Yaffee, M. (1995) Biochim. Biophys. Acta 1271, 67–74
 Radical-mediated protein oxidation 17

159 Dean, R. T. and Pollak, J. K. (1985) Biochem. Biophys. Res. Commun. 126, 207 Aguirre, J., Rodriguez, R. and Hansberg, W. (1989) J. Bacteriol. 171, 6243–6250
1082–1089 208 Aupeix, K., Toti, F., Satta, N., Bischoff, P. and Freyssinet, J. M. (1996) Biochem. J.
160 Marcillat, O., Zhang, Y., Lin, S. W. and Davies, K. J. (1988) Biochem. J. 254, 314, 1027–1033
677–683 209 Bjorkerud, B. and Bjorkerud, S. (1996) Arterioscler. Thromb. Vasc. Biol. 16,
161 Ohad, I., Kyle, D. J. and Arntzen, C. J. (1984) J. Cell Biol. 99, 481–485 416–424
162 Keren, N., Gong, H. and Ohad, I. (1995) J. Biol. Chem. 270, 806–814 210 Ho, Y. S., Wang, Y. J. and Lin, J. K. (1996) Mol. Carcinog. 16, 20–31
163 Masuda, Y. and Murano, T. (1979) Jpn. J. Pharmacol. 29, 179–186 211 Rosselli, F., Ridet, A., Soussi, T., Duchaud, E., Alapetite, C. and Moustacchi, E.
164 Ando, M. and Tappel, A. L. (1985) Chem.-Biol. Interact. 55, 317–326 (1995) Oncogene 10, 9–17
165 Mukhopadhyay, M., Mukhopadhyay, C. K. and Chatterjee, I. B. (1993) Mol. Cell. 212 Selvaraj, R. J., Paul, B. B., Strauss, R. R., Jacobs, A. A. and Sbarra, A. J. (1974)
Biochem. 126, 69–75 Infect. Immun. 9, 255–260
166 Mukhopadhyay, C. K., Ghosh, M. K. and Chatterjee, I. B. (1995) Mol. Cell. Biochem. 213 Aune, T. M. and Thomas, E. L. (1978) Biochemistry 17, 1005–1010
142, 71–78 214 Thomas, E. L. and Aune, T. M. (1978) Antimicrob. Agents Chemother. 13,
167 Dean, R. T., Roberts, C. R. and Forni, L. G. (1984) Biosci. Rep. 4, 1017–1026 1006–1010
168 Roberts, C. R., Roughley, P. J. and Mort, J. S. (1989) Biochem. J. 259, 805–811 215 Thomas, E. L., Jefferson, M. M. and Grisham, M. B. (1982) Biochemistry 21,
169 Chung, M. H., Kesner, L. and Chan, P. C. (1984) Agents Actions 15, 328–335 6299–6308
170 Grant, A. J., Jessup, W. and Dean, R. T. (1993) Free Radical Res. Commun. 18, 216 Roberts, C. R. and Dean, R. T. (1986) Connect. Tissue Res. 14, 199–212
259–267 217 Vissers, M. C. M. and Winterbourn, C. C. (1986) Biochim. Biophys. Acta 889,
171 Jessup, W., Mander, E. L. and Dean, R. T. (1992) Biochim. Biophys. Acta 1126, 277–286
167–177 218 Weiss, S. J., Curnutte, J. T. and Regiani, S. (1986) J. Immunol. 136, 636–641
172 Ballard, F. J. and Hopgood, M. F. (1976) Biochem. J. 154, 717–724 219 Mukhopadhyay, C. K. and Chatterjee, I. B. (1994) J. Biol. Chem. 269,
173 Francis, G. L. and Ballard, F. J. (1980) Biochem. J. 186, 571–579 30200–30205
174 Toledo, I. and Hansberg, W. (1990) Exp. Mycol. 14, 184–189 220 Stief, T. W., Stief, M. H., Ehrenthal, W., Darius, H. and Martin, E. (1991) Thromb.
175 Hansberg, W., De Groot, H. and Sies, H. (1993) Free Radical Biol. Med. 14, Res. 64, 597–610
287–293 221 Costabel, U., Maier, K., Teschler, H. and Wang, Y. M. (1992) Respiration 1, 17–19
176 Toledo, I., Aguirre, J. and Hansberg, W. (1994) Microbiology 140, 2391–2397 222 Oliver, C. N. (1987) Arch. Biochem. Biophys. 253, 62–72
177 Toledo, I., Rangel, P. and Hansberg, W. (1995) Arch. Biochem. Biophys. 319, 223 Kettle, A. J. (1996) FEBS Lett. 379, 103–106
519–524 224 Richards, D. M., Dean, R. T. and Jessup, W. (1988) Biochim. Biophys. Acta 946,
178 Dean, R. T. (1975) Biochem. Biophys. Res. Commun. 67, 604–609 281–288
179 Dean, R. T. (1975) Nature (London) 257, 414–416 225 Yamada, M., Hearse, D. J. and Curtis, M. J. (1990) Circ. Res. 67, 1211–1224
180 Cuervo, A. M., Knecht, E., Terlecky, S. R. and Dice, J. F. (1995) Am. J. Physiol. 226 Kusama, Y., Bernier, M. and Hearse, D. J. (1989) Circulation 80, 1432–1448
269, C1200–C1208 227 Bernier, M., Kusama, Y., Borgers, M., Ver, D. L., Valdes, A. O., Neckers, D. C. and
181 Wiertz, E. J. H. J., Tortorella, D., Bogyo, M., Yu, J., Mothes, W., Jones, T. R., Hearse, D. J. (1991) Free Radical Biol. Med. 10, 287–296
Rapoport, T. A. and Ploegh, H. L. (1996) Nature (London) 384, 432–438
228 Yasutake, M., Ibuki, C., Hearse, D. J. and Avkiran, M. (1994) Am. J. Physiol. 267,
182 Mander, E. L., Dean, R. T., Stanley, K. K. and Jessup, W. (1994) Biochim. Biophys.
H2430–H2440
Acta 1212, 80–92
229 Coetzee, W. A., Ichikawa, H. and Hearse, D. J. (1994) Am. J. Physiol. 266,
183 Grant, A. J., Jessup, W. and Dean, R. T. (1992) Biochim. Biophys. Acta 1134,
H909–H919
203–209
230 Chevion, M., Jiang, Y., Har, E. R., Berenshtein, E., Uretzky, G. and Kitrossky, N.
184 Matthews, W., Driscoll, J., Tanaka, K., Ichihara, A. and Goldberg, A. L. (1989) Proc.
(1993) Proc. Natl. Acad. Sci. U.S.A. 90, 1102–1106
Natl. Acad. Sci. U.S.A. 86, 2597–2601
231 Garlick, P. B., Davies, M. J., Hearse, D. J. and Slater, T. F. (1987) Circ. Res. 61,
185 Grune, T., Reinheckel, T., Joshi, M. and Davies, K. J. (1995) J. Biol. Chem. 270,
757–760
2344–2351
232 Poston, J. M. and Parenteau, G. L. (1992) Arch. Biochem. Biophys. 295, 35–41
186 Hendil, K. B. (1980) J. Cell. Physiol. 105, 449–460
187 Dean, R. T., Armstrong, S. G., Fu, S. and Jessup, W. (1994) in Free Radicals in the 233 Liu, Y., Rosenthal, R. E., Starke, R. P. and Fiskum, G. (1993) Free Radical Biol.
Environment, Medicine and Toxicology (Nohl, H., Esterbauer, H. and Rice-Evans, C., Med. 15, 667–670
eds.), pp. 47–79, Richelieu Press, London 234 Carney, J. M. and Floyd, R. A. (1991) J. Mol. Neurosci. 3, 47–57
188 Gershon, H. and Gershon, D. (1970) Nature (London) 227, 1214–1216 235 Krause, G. S., DeGracia, D. J., Skjaerlund, J. M. and O’Neil, B. J. (1992)
189 Cao, G. and Cutler, R. G. (1995) Arch. Biochem. Biophys. 320, 106–114 Resuscitation 23, 59–69
190 Levine, R. L., Williams, J. A., Stadtman, E. R. and Shacter, E. (1994) Methods 236 Bogaert, Y. E., Rosenthal, R. E. and Fiskum, G. (1994) Free Radical Biol. Med. 16,
Enzymol. 233, 346–357 811–820
191 Lyras, L., Evans, P. J., Shaw, P. J., Ince, P. G. and Halliwell, B. (1996) Free Radical 237 Brunk, U. T., Jones, C. B. and Sohal, R. S. (1992) Mutat. Res. 275, 395–403
Res. Commun. 24, 397–406 238 Stadtman, E. R., Starke-Reed, P. E., Oliver, C. N., Carney, J. M. and Floyd, R. A.
192 Marquez, L. A. and Dunford, H. B. (1995) J. Biol. Chem. 270, 30434–30440 (1992) EXS (Basel) 62, 64–72
193 Wilson, J. B., Brennan, S. O., Allen, J., Shaw, J. G., Gu, L. H. and Huisman, T. H. 239 Carney, J. M., Starke, R. P., Oliver, C. N., Landum, R. W., Cheng, M. S., Wu, J. F.
(1993) J. Chromatogr. 617, 37–42 and Floyd, R. A. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 3633–3636
194 Rzepecki, L. M. and Waite, J. H. (1995) Mol. Mar. Biol. Biotechnol. 4, 313–322 240 Dubey, A., Forster, M. J. and Sohal, R. S. (1995) Arch. Biochem. Biophys. 324,
195 Taylor, S. W., Molinski, T. F., Rzepecki, L. M. and Waite, J. H. (1991) J. Nat. Prod. 249–254
54, 918–922 241 Garner, M. H. and Spector, A. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 1274–1277
196 Rzepecki, L. M. and Waite, J. H. (1993) Mol. Mar. Biol. Biotechnol. 2, 255–266 242 Wells-Knecht, M. C., Huggins, T. G., Dyer, D. G., Thorpe, S. R. and Baynes, J. W.
197 Abe, M. K., Chao, T. S., Solway, J., Rosner, M. R. and Hershenson, M. B. (1994) (1993) J. Biol. Chem. 268, 12348–12352
Am. J. Respir. Cell. Mol. Biol. 11, 577–585 243 Sohal, R. S., Agarwal, S. and Sohal, B. H. (1995) Mech. Ageing Dev. 81, 15–25
198 Storz, G. and Tartaglia, L. A. (1992) J. Nutr. 122, 627–630 244 Cini, M. and Moretti, A. (1995) Neurobiol. Aging 16, 53–57
199 Barchowsky, A., Munro, S. R., Morana, S. J., Vincenti, M. P. and Treadwell, M. 245 Stadtman, E. R. (1992) Science 257, 1220–1224
(1995) Am. J. Physiol. 269, L829–L836 246 Sohal, R. S. and Dubey, A. (1994) Free Radical Biol. Med. 16, 621–626
200 Goldstone, S. D., Fragonas, J. C., Jeitner, T. M. and Hunt, N. H. (1995) Biochim. 247 Orr, W. C. and Sohal, R. S. (1994) Science 263, 1128–1130
Biophys. Acta 1263, 114–122 248 Sohal, R. S., Agarwal, A., Agarwal, S. and Orr, W. C. (1995) J. Biol. Chem. 270,
201 Chen, Q., Fischer, A., Reagan, J. D., Yan, L. J. and Ames, B. N. (1995) Proc. Natl. 15671–15674
Acad. Sci. U.S.A. 92, 4337–4341 249 Sohal, R. S., Arnold, L. and Orr, W. C. (1990) Mech. Ageing Dev. 56, 223–235
202 Oliver, C. N., Ahn, B. W., Moerman, E. J., Goldstein, S. and Stadtman, E. R. (1987) 250 Sohal, R. S. and Weindruch, R. (1996) Science 273, 59–63
J. Biol. Chem. 262, 5488–5491 251 Wolff, S. P. and Dean, R. T. (1987) Biochem. J. 245, 243–250
203 Fulks, R. M. and Stadtman, E. R. (1985) Biochim. Biophys. Acta 843, 214–229 252 Wolff, S. P. and Dean, R. T. (1987) Bioelectrochem. Bioenerg. 18, 283–293
204 Hansberg, W. and Aguirre, J. (1990) J. Theor. Biol. 142, 201–221 253 Hicks, M., Delbridge, L., Yue, D. K. and Reeve, T. S. (1988) Biochem. Biophys. Res.
205 Toledo, I., Noronha-Dutra, A. A. and Hansberg, W. (1991) J. Bacteriol. 173, Commun. 151, 649–655
3243–3249 254 Chace, K. V., Carubelli, R. and Nordqvist, R. E. (1991) Arch. Biochem. Biophys.
206 Hansberg, W. (1996) Cienc. Cult. (Sao Paulo) 48, 68–74 288, 473–480
18 R. T. Dean and others

255 Dean, R. T., Gebicki, J., Gieseg, S., Grant, A. J. and Simpson, J. A. (1992) Mutat. 283 Stadtman, E. R. (1993) Annu. Rev. Biochem. 62, 797–821
Res. 275, 387–393 284 Kim, K., Kim, I. H., Lee, K. Y., Rhee, S. G. and Stadtman, E. R. (1988) J. Biol.
256 Dunn, J. A., McCance, D. R., Thorpe, S. R., Lyons, T. J. and Baynes, J. W. (1991) Chem. 263, 4704–4711
Biochemistry 30, 1205–1210 285 Rhee, S. G., Kim, K. H., Chae, H. Z., Yim, M. B., Uchida, K., Netto, L. E. and
257 Sell, D. R. and Monnier, V. M. (1989) J. Biol. Chem. 24, 21597–21602 Stadtman, E. R. (1994) Ann. N. Y. Acad. Sci. 738, 86–92
258 Suarna, C., Dean, R. T., May, J. and Stocker, R. (1995) Arterioscler. Thromb. Vasc. 286 Forsmark, A. P., Dallner, G. and Ernster, L. (1995) Free Radical Biol. Med. 19,
Biol. 15, 1616–1624 749–757
259 Yla-Herttuala, S., Luoma, J., Viita, H., Hiltunen, T., Sisto, T. and Nikkari, T. (1995) 287 Gebicki, S., Bartosz, G. and Gebicki, J. M. (1995) Biochem. Soc. Trans. 23, 249S
J. Clin. Invest. 95, 2692–2698 288 Sattler, W., Maiorino, M. and Stocker, R. (1994) Arch. Biochem. Biophys. 309,
260 Beckman, J. S., Ye, Y. Z., Anderson, P. G., Chen, J., Accavitti, M. A., Tarpey, M. M. 214–221
and White, C. R. (1994) Biol. Chem. Hoppe-Seyler 375, 81–88 289 Christison, J., Sies, H. and Stocker, R. (1994) Biochem. J. 304, 341–345.
261 Uchida, K., Itakura, K., Kawakishi, S., Hiai, H., Toyokuni, S. and Stadtman, E. R. 290 Chae, H. Z., Chung, S. J. and Rhee, S. G. (1994) J. Biol. Chem. 269, 27670–27678
(1995) Arch. Biochem. Biophys. 324, 241–248 291 Bjornstedt, M., Hamberg, M., Kumar, S., Sue, J. and Homgren, A. (1995) J. Biol.
262 Hazell, L. J., Arnold, L., Flowers, D., Waeg, G., Malle, E. and Stocker, R. (1996) J. Chem. 270, 11761–11764
Clin. Invest. 97, 1535–1544 292 Yim, M. B., Chae, H. Z., Rhee, S. G., Chock, P. B. and Stadtman, E. R. (1994) J.
263 Hazell, L. J. and Stocker, R. (1993) Biochem. J. 290, 165–172 Biol. Chem. 269, 1621–1626
264 Dabbagh, A. J. and Frei, B. (1995) J. Clin. Invest. 96, 1958–1966 293 Neuzil, J., Thomas, S. R. and Stocker, R. (1997) Free Radical Biol. Med. 22, 57–71
265 Stocker, R. (1994) Curr. Opin. Lipidol. 5, 422–433 294 van der Vliet, A., Eiserich, J. P., O’Neill, C. A., Halliwell, B. and Cross, C. E. (1995)
266 Bowry, V. W., Mohr, D., Cleary, J. and Stocker, R. (1995) J. Biol. Chem. 270, Arch. Biochem. Biophys. 319, 341–349
5756–5763 295 Huggins, T. G., Wells, K. M., Detorie, N. A., Baynes, J. W. and Thorpe, S. R. (1993)
267 Parhami, F., Fang, Z. T., Fogelman, A. M., Andalibi, A., Territo, M. C. and Berliner, J. Biol. Chem. 268, 12341–12347
J. A. (1993) J. Clin. Invest. 92, 471–478 296 Armstrong, R. C. and Swallow, A. J. (1969) Radiat. Res. 40, 563–579
297 Maskos, Z., Rush, J. D. and Koppenol, W. H. (1992) Arch. Biochem. Biophys. 262,
268 Smith, C. D., Carney, J. M., Tatsumo, T., Stadtman, E. R., Floyd, R. A. and
521–529
Markesbery, W. R. (1992) Ann. N. Y. Acad. Sci. 663, 110–119
298 Stadtman, E. R. and Berlett, B. S. (1991) J. Biol. Chem. 266, 17201–17211
269 Montine, T. J., Amarnath, V., Martin, M. E., Strittmatter, W. J. and Graham, D. G.
299 Trelstad, R. L., Lawley, K. R. and Holmes, L. B. (1981) Nature (London) 289,
(1996) Am. J. Pathol. 148, 89–93
310–322
270 Dyrks, T., Dyrks, E., Masters, C. L. and Beyreuther, K. (1993) FEBS Lett. 324,
300 Ayala, A. and Cutler, R. G. (1996) Free Radical Biol. Med. 21, 65–80
231–236 301 Wheelan, P., Kirsch, W. M. and Koch, T. H. (1989) J. Org. Chem. 54, 4360–4364
271 Dykens, J. A. (1994) J. Neurochem. 63, 584–591 302 Copley, S. D., Frank, E., Kirsch, W. M. and Koch, T. H. (1992) Anal. Biochem. 201,
272 Smith, C. D., Carney, J. M., Starke, R. P., Oliver, C. N., Stadtman, E. R., Floyd, R. A. 152–157
and Markesbery, W. R. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 10540–10543 303 Li, S., Schoneich, C. and Borchardt, R. T. (1995) Pharmacol. Res. 12, 348–355
273 Smith, M. A., Perry, G., Richey, P. L., Sayre, L. M., Anderson, V. E., Beal, M. F. and 304 Li, S., Nguyen, T. H., Schoneich, C. and Borchardt, R. T. (1995) Biochemistry 34,
Kowall, N. (1996) Nature (London) 382, 120–121 5762–5772
274 Smith, M. A., Richey, P. L., Taneda, S., Kutty, R. K., Sayre, L. M., Monnier, V. M. 305 Armstrong, D. A. (1990) in Sulfur-Centered Reactive Intermediates in Chemistry and
and Perry, G. (1994) Ann. N. Y. Acad. Sci. 738, 447–454 Biology (Chatgilialoglu, C. and Asmus, K.-D., eds.), pp. 121–134, Plenum Press,
275 Forster, M. J., Dubey, A., Dawson, K. M., Stutts, W. A., Lal, H. and Sohal, R. S. New York
(1996) Proc. Natl. Acad. Sci. U.S.A. 93, 4765–4769 306 von Sonntag, C. (1990) in Sulfur-Centered Reactive Intermediates in Chemistry and
276 Floyd, R. A. and Carney, J. M. (1992) Ann. Neurol. 32, S22–S27 Biology (Chatgilialoglu, C. and Asmus, K.-D., eds.), pp. 359–366, Plenum Press,
277 Puttfarcken, P. S., Manelli, A. M., Neilly, J. and Frail, D. E. (1996) Exp. Neurol. New York
138, 73–81 307 Rowbottom, J. (1955) J. Biol. Chem. 212, 877–885
278 Hensley, K., Carney, J. M., Mattson, M. P., Aksenova, M., Harris, M., Wu, J. F., 308 Nosworthy, J. and Allsop, C. B. (1956) J. Colloid Sci. 11, 565–574
Floyd, R. A. and Butterfield, D. A. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 309 Hazen, S. L., Hsu, F. F. and Heinecke, J. W. (1996) J. Biol. Chem. 271,
3270–3274 1861–1867
279 Harris, M. E., Hensley, K., Butterfield, D. A., Leedle, R. A. and Carney, J. M. (1995) 310 Leeuwenburgh, C., Rasmussen, J. E., Hsu, F. F., Mueller, D. M., Pennathur, S. and
Exp. Neurol. 131, 193–202 Heinecke, J. W. (1997) J. Biol. Chem. 272, 3520–3526
280 Benzi, G. and Moretti, A. (1995) Neurobiol. Aging 16, 661–674 311 Van Buskirk, J. J., Kirsch, W. M., Kleyer, D. L., Darklen, R. M. and Koch, T. H.
281 Pellmar, T. C. (1995) J. Neurosci. Methods 59, 93–98 (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 722–725
282 Simpson, J. A. and Dean, R. T. (1990) Free Radical Res. Commun. 10, 303–312 312 Ayala, A. and Cutler, R. G. (1996) Free Radical Biol. Med. 21, 551–558