[Advanced Topics in Science and Technology in China] Selenoproteins and Mimics Volume 703 || Glutathione Peroxidases

1

Glutathione Peroxidases

Leopold Flohé

Otto-von-Guericke-Universität, Universitätsplatz 2, D-39106 Magdeburg, and

MOLISA GmbH, Brenneckestrasse 20, D-39118 Magdeburg, Germany

E-mail: [email protected]

The present state of knowledge on glutathione peroxidases (GPxs) is reviewed

with particular emphasis on general catalytic principles and the biology of

mammalian glutathione peroxidases. GPxs make up a ubiquitous family of

proteins defined by sequence homology, the common functional denominator

being their ability to reduce hydroperoxides by thiols. Catalysis is mediated by an

active-site selenocysteine or cysteine. Eight distinct GPxs have been identified in

mammals, five of them being selenoproteins in man. While glutathione specificity

prevails in vertebrate GPxs, thioredoxins or related redoxins appear to be common

substrates in plant, bacterial and protist GPxs. Specific reactions of GPxs with

other protein thiols are also observed. The basic catalytic scheme allows the

enzymes to adopt diversified biological roles ranging from defence against

peroxide challenge, redox regulation of metabolic processes and transcription,

apoptosis to cellular differentiation. The roles of the individual mammalian GPxs

are discussed in the light of distinct substrate specificities, distribution, subcellular

compartmentation, expression patterns and data from inverse genetics. It is

outlined that the multiple coexisting GPxs and functionally related peroxiredoxins

likely build up a system of enzymes that, with discrete functional overlap,

complement each other in meeting specific biological tasks far beyond fighting

oxidative stress.

1.1� Introduction

In 1973, glutathione peroxidase (GPx) was identified as a selenoprotein, in fact the

J. Liu et al. , Selenoproteins and Mimics© Zhejiang University Press, Hangzhou and Springer-Verlag Berlin Heidelberg 2011

(eds.)

1� Glutathione Peroxidases

2

first one to be discovered in higher organisms [1, 2]

. The enzyme, which is now

known as GPx1 catalyzed the reduction of H2O

2 and organic hydroperoxides by

glutathione. Its selenoprotein nature finally explained why traces of selenium are

essential for defense against an oxidative challenge in the vertebrate organism.

The fact that the first mammalian selenoprotein was a peroxidase, however, also

led to the misconception that the essential trace element selenium is simply a

“biological antioxidant”. The early history of glutathione peroxidase and selenium

biochemistry with all its interdependencies, serendipities and surprises were the

subject of a recent essay [3]

and shall not be repeated here in detail. It may suffice

to say that research on GPxs was pivotal to our present understanding of

selenium’s role in biology, its function as a catalytic entity in enzymes, as well as

providing an understanding of a most complex mechanism in its co-translational

incorporation into selenoproteins [4-8]

.

Over the past three decades, GPxs, defined as proteins with high sequence

similarity, have been detected in almost every domain of life. The majority of

these proteins, however, are neither selenoproteins nor glutathione peroxidases, if

this term is to characterize their catalytic role. The selenium-containing

glutathione peroxidases prevail in vertebrates and have only been sporadically

detected in lower organisms such as platyhelminths (e.g. Schistosoma mansoni, S.

japonicum and Echinococcus granulosus) [9, 10]

, Cnidaria (Hydra vulgaris) and

protists [11]

, and exceptionally in insects (in the tick Boophilus microplus) [12]

and

bacteria [13]

. In most of the invertebrate species, all yeasts and higher plants, the

active site selenocysteine of the glutathione peroxidases is replaced by cysteine.

Interestingly, this change in the redox-active moiety is often associated with a

switch in substrate specificity: most of the non-Se glutathione peroxidases appear

to hardly react with glutathione (GSH). Instead, these GPx homologues, like most

of the peroxiredoxins, are preferentially or exclusively reduced by “redoxin”-type

proteins such as thioredoxin [14]

or tryparedoxin [15]

. A yeast GPx homologue has

also been described to specifically react with a particular SH group of a

transcription factor and to thereby initiate the expression of protective enzymes [16]

.

However, the selenium-containing GPxs are not always specific for GSH either. In

fact, a strict specificity for GSH has only been documented for the prototype

which gave the name to the entire family, i.e. GPx1 [17]

, whereas e.g. GPx4 has

been reported to react with a variety of protein thiols [18-20]

including SH groups of

GPx4 itself [21, 22]

.

The ramification that the GPx family experienced during evolution [11, 23]

renders it obsolete to talk about “glutathione peroxidase” as a functionally well-

defined enzymatic entity. Many of the family members might not at all share the

basic biological role of GPx1, which is to reduce H2O

2 or other hydroperoxides at

the expense of GSH to cope with oxidative challenge. Moreover, the term

“glutathione peroxidase” has been used to describe enzymes that may similarly

catalyze the reduction of hydroperoxides by GSH, but are neither structurally nor

phylogenetically related to the family, such as GSH-S-transferases [24]

, selenoprotein

P [25]

or human peroxiredoxin VI [26]

.

The growing complexity of thiol-dependent hydroperoxide metabolism has

1.2� Glutathione Peroxidase Reaction

3

been discussed in many topical reviews, each one focusing on particular aspects

such as evolution [11, 23]

, specificities [20]

, kinetics [27, 28]

, catalytic mechanism [20, 27]

,

regulation of enzyme expression [29, 30]

and its involvement in redox regulation [30-32]

,

male fertility [33, 34]

, apoptosis [35, 36]

, viral infections [37]

, thyroid [38]

or brain function [39]

.

And the overlap between the glutathione- and thioredoxin-dependent hydroperoxide

metabolizing systems may be distilled from respective monographs [40, 41]

. By the

end of June 2009, a PubMed search for the key word “glutathione peroxidase”

yielded 10,928 entries, which reveals the impossibility of covering the entire field

in this review with an allotted maximum length of 20 pages. This article will

therefore be essentially confined to general aspects of GPx catalysis and the

peculiarities of the mammalian selenium-containing peroxidases.


The GPx that gave its name to the entire family [3]

catalyzes the reduction of H2O

2

and soluble organic hydroperoxides at the expense of GSH. This first glutathione

peroxidase, now called GPx1, is a tetrameric enzyme consisting of four identical

subunits (Fig. 1.1). Having remained the only known GPx for more than two

decades, it also served as a prototype for working out the kinetic mechanism,

sequence and structure, specificity and the catalytic principle which involves

oxidation of the active site selenium and step-wise reduction by GSH. The present

mechanistic understanding of this enzyme, which is widely relevant to other types

of GPx, is critically reviewed in the paragraphs below.

Fig. 1.1.� Structure of GPx1. The representation shows the homo-tetrameric enzyme with its four

selenium atoms as orange balls. Reproduced from the data set of Epp et al. [52]

by K. D. Aumann,

Helmholtz-Zentrum für Infektionsforschung, Braunschweig, Germany


4

1.2.1� Basic Catalytic Principle

Typically, glutathione peroxidases catalyze the reduction of H2O

2 by GSH

according to Eq.(1.1).

H2O

2 + 2 GSH → GSSG + 2 H

2O (1.1)

Depending on the particular enzyme, a more or less broad scope of

hydroperoxides may be reduced (Eq.(1.2)) ,

ROOH + 2 GSH → GSSG + ROH + H2O (1.2)

and the reductant GSH may be partially or fully replaced by other thiols (Eq.(1.3)),

H2O

2 + GSH + RSH → GSSR + 2 H

2O (1.3)

or

H2O

2 + 2 RSH → RSSR + 2 H

2O (1.4)

All these reactions seem to be chemically trivial and indeed proceed

spontaneously, provided the thiol groups are dissociated. In reality, however, these

reactions (Eqs.(1.1) – (1.4)), each one involving three molecules, proceed according

to a lower order of kinetics than anticipated, since they do not require any ternary

collision of the three molecules but result from a sequence of two binary collisions

(Eqs.(1.5) and (1.6)). The first thiol, which has to be present in its thiolate form,

reacts with the hydroperoxide, whereby a sulfenic acid is formed.

RS−

+ H2O

2→ RSOH + OH

−

(1.5)

The latter then dissociates and reacts with the second thiol to form the disulfide.

RSO−

+ H+

+RSH → RSSR + H2O (1.6)

The enzymatic catalysis of hydroperoxide reduction by thiols mimics much of

this simple chemistry but speeds up the reaction rate by orders of magnitude

(Eqs.(1.7) and (1.8)). The enzymatic trick is that the hydroperoxide has not to

directly attack a poorly reactive, since hardly dissociated, thiol such as GSH.

Instead, a highly-reactive thiol or selenol within the enzyme, which is the

“peroxidatic cysteine” (CP) or “peroxidatic selenocysteine” (U

P), reduces the

hydroperoxide.

E-S−

+ H2O

2 → E-SOH + OH

−

(1.7)

E-Se−

+ H2O

2 → E-SeOH + OH

−

(1.8)

The sulfenic or selenenic acid residue of the enzyme, in analogy to Eq.(1.6),

readily reacts with the substrate thiol to form a mixed (selena) disulfide which, by

thiol-disulfide exchange, is reduced by the second thiol (Eqs.(1.9) and (1.10)).

E-S(Se)O−

+ H+

+ RSH → E-S(Se)SR + H2O (1.9)


5

E-S(Se)SR + RSH → E-S(Se)−

+ H+

+ RSSR (1.10)

In essence this basic scheme is valid for the “real” glutathione peroxidases and,

with some modifications, also for the GPx-type [14]

and peroxiredoxin-type

thioredoxin peroxidases [41]

. In the latter two cases, the first reducing thiol is a

cysteine residue of the enzyme itself (called the “resolving” cysteine, CR), and the

resulting disulfide form of the enzyme is then reduced by the CXXC motif of a

redoxin-type protein.

1.2.2� Kinetics

The catalytic principle outlined above is best described by the German term

“Zwischenstoffkatalyse” (catalysis by intermediate formation), as it was developed

by the German chemist Wilhelm Ostwald in the beginning of the last century [42]

.

It means that the catalysis is achieved by a sequence of partial reactions of the

reactants with the catalyst and modifications thereof, each being faster than the

non-catalyzed overall reaction. In our example, the catalyst is oxidized by H2O

2

(Eq.(1.8)), and the intermediate E-SeOH (Zwischenstoff) thus formed is stepwise

reduced by GSH (Eqs.(1.9) and (1.10)). The correlation in enzymological terms is

the “enzyme substitution mechanism”, as defined by Dalziel in 1957 [43]

. This

catalytic principle, which is by no means uncommon for oxidoreductases, differs

substantially from “central complex mechanisms”, where two or more reactants

are assembled at the enzyme’s active site in a productive way to facilitate their

interaction. This difference between mechanisms has to be stressed, because it has

a major impact on kinetics and, in consequence, on the enzyme’s function in a

biological context. While enzymes with central complex mechanisms are best

characterized by Michaelis constants and maximum velocities, these classical

parameters adopt a completely different physical meaning in enzyme substitution

mechanisms or, as in the case of the selenoperoxidases, may not be applicable at

all: The Km and V

max values of GPx1

[44]

and all other selenium-containing GPxs

investigated so far are infinite [27]

. This seemingly odd behavior does not reflect a

low affinity of substrates to GPx, but simply reveals a high reactivity of the

substrates with the enzyme or its derivates, respectively. In contrast to the

Michaelis-Menten theorem, it is not a reaction of substrates within an enzyme /

substrate complex that is rate-limiting in the GPx reaction, but the speed of

productive collisions of the ground-state enzyme with a hydroperoxide (Eq.(1.8))

or the formation of binary complexes between GSH and one of the oxidized

enzyme forms (Eqs.(1.9) and (1.10)). Although the formation of such complexes

(omitted in Eqs.(1.9) and (1.10)) is not evidenced by steady-state kinetics, they

have to be inferred for GPx1 at least from its pronounced donor substrate

specificity. In line with this interpretation, saturation kinetics are sometimes

observed with GPx homologues working with the less reactive CP [27]

or with other

thiol peroxidases relying on sulphur catalysis [28]

.


6

With the above consideration, the initial rate equation for GPx1 [44]

and

identically for GPx3 [45]

and GPx4 [45, 46]

becomes surprisingly simple (Eq.(1.11)):

[E0] / v

0 = 1 / k

+́1·[ROOH] + 1 / k

+́2 ·[GSH] (1.11)

Therein k+́1

is the apparent net forward rate constant for partial reaction (Eq.(1.8))

and, in view of the irreversibility of this step and lacking evidence for a specific

enzyme/hydroperoxide complex, may be regarded as the bimolecular rate constant

k+1

that characterizes the oxidation of the ground state enzyme with the

hydroperoxide. k+́2

is less well defined. It is the net forward rate constant for the

reductive part of the catalytic cycle and physically means the net forward rate

constant for the association of GSH with the oxidized (Eq.(1.12)) or partially

reduced enzyme (Eq.(1.13)), whichever is smaller, or a hybrid constant, if they are

similar.

E-SeOH + GSH → [E-SeOH·GSH] (1.12)

E-SeSG + GSH → [E-SeSG·GSH] (1.13)

The complexes, however, never accumulate and therefore remain kinetically

silent, since the reactions according to Eq.(1.9) and Eq.(1.10) proceed within these

complexes with a non-rate-limiting, i.e. higher velocity. Despite its poorly-defined

physical meaning, k+́2

is a useful constant to predict turnover rates under varying

physiological conditions.

1.2.3� Physiological Consequences of Kinetic Mechanism

For all mammalian selenium-containing GPxs so far analyzed, a k+1

> 107

L/(mol·s)

(for H2O

2 ) was determined, whereas the k

+́2 is two to three orders of magnitude

smaller. As the oxidative step is so much faster than the reductive ones, the

enzyme is almost 100% oxidized if its velocity is measured at similar substrate

concentrations, as commonly done in vitro. Under such conditions the rate

equation (Eq.(1.11)) simplifies to Eq.(1.14):

v0 = k

+́2 · [GSH] · [E

0] (1.14)

which means that the turnover depends on the concentration of GSH and over a

wide range is independent of the H2O

2 concentrations. In fact, the enzyme seems

always “saturated” with H2O

2 and an apparent K

M is hard to measure. This

observation has frequently led to the misconception that the enzymes similarly

respond to variations in substrate concentrations in vivo. The opposite is correct:

the general rate equation (Eq.(1.12)) yields that, at physiological substrate

concentrations of 1 – 10 mmol/L GSH and an estimated maximum of 1 μmol/L

H2O

2 or other hydroperoxides, the enzyme is largely reduced, even if k

+1 is two

orders of magnitude larger than k+́2

. With [E0] = [E

red], however, the rate equation

simplifies to Eq.(1.15).


7

v0 = k

+́1 · [ROOH] [E

0] (1.15)

This implies that in vivo the GPx turnover in most cells is independent of the

concentration of GSH, unless it drops to less than 10−4

mol/L. This straightforward

consequence of the kinetic parameters of the enzymes seemingly conflicts with

observations relating impaired antioxidant defense to moderately-lowered GSH

content in tissues. The solution of the enigma is provided by uneven GSH

concentrations in cells and cellular compartments. A drop in GSH by, e.g. 20%,

likely means that GSH is practically zero in 20% of the cells. Such GSH depletion

is not reached before the rate of H2O

2 production exceeds the rate of GSH

regeneration by glutathione reductase or the NADPH supply systems, respectively.

Exceptionally, this happens physiologically in special cells but commonly marks a

transition point from physiology to pathophysiology.

1.2.4� Facts, Unknowns and Guesswork

While the basic principles of the glutathione peroxidase reaction, according to

Eqs.(1.7) – (1.13), are generally accepted, many details still remain enigmatic.

1.2.4.1� Catalytic Relevance of UP or C

P Dissociation

From alkylation studies and general chemical considerations, it appears obvious

that the ground state enzyme presents its active site selenocysteine as a selenolate,

and the extreme efficiency of the selenoperoxidases is usually explained by the

comparatively low pKa of selenocysteine (pK

a = 5.2) versus cysteine (pK

a = 8.3).

Unfortunately, this reasoning, although repeated even in most recent publications [47]

,

does not really lead to any satisfactory interpretation of experimental data:

i) Fully-dissociated low molecular weight thiols do not react with H2O

2 faster

than with bimolecular rate constants near 50 L/(mol·s) [48]

, while corresponding

rate constants for cysteine residues in GPx- or Prx-type peroxidases ranging

around 106

L/(mol·s) are by no means exceptional [27, 28]

.

ii) Within the architecture of the GPx active site CP or U

P appear to be similarly

dissociated, as has been suggested by pKa calculations

[20]

and demonstrated by

velocities of alkylation [49, 50]

, which equally requires the thiolate or selenolate

form, respectively. The efficiencies of recombinant cysteine homologues of Se-

GPxs, however, are typically three orders of magnitude smaller [49-51]

and k+1

values

near 108

L/(mol·s), as determined for natural GPx1 [44]

, have never been observed

with any of the thiol peroxidases working with sulphur catalysis [27, 28]

.

iii) The electro-negativity of sulphur and selenium does not differ significantly

enough to account for the substantial difference in catalytic efficiency either. Thus,

in short, the dissociation of CP or U

P, respectively, although being a prerequisite


8

for the enzyme’s reaction with ROOH, neither explains the catalytic efficiency of

GPxs in general nor the superiority of the selenium-containing ones.

1.2.4.2� Mechanism of UP Activation

An activation of the UP by neighboring residues had already been deduced from

the first X-ray structure of a GPx, that of GPx1 [52]

. Although the UP in this

structure was over-oxidized to a seleninic acid, it seemed plausible that in the

ground state enzyme the selenium atom might be hydrogen-bonded to the amide

nitrogen of a glutamine and the imino nitrogen of a tryptophan, whereby a

catalytic triad consisting of Sec (or Cys), Gln and Trp is formed, which over the

years became a characteristic signature of the entire family. In this triad the

selenol function should be forced into dissociation and further polarized for a

nucleophilic attack on the peroxy bond of the substrate. The catalytic relevance of

these conserved residues could indeed be verified by site-directed mutagenesis of

GPx4 [49]

and others [15, 50]

. More recently, the triad concept had to be amended,

since a strictly conserved Asn that contacts the UP or C

P from the core of the

protein proved to have an even higher impact on activity than Gln and Trp so far

implicated, whereby the catalytic triad grew up to a tetrad [20]

. The residues were

shown to facilitate dissociation of CP (a U

P would be dissociated anyway)

[20, 49, 50]

and S-alkylation [49, 50]

and thus contribute to the nucleophilicity of the active site

chalcogen, which is a necessary, though not sufficient, condition for catalytic

efficiency (see subsection 3.4.1). Recent re-calculations based on all known GPx

structures, however, revealed that these residues are in an ideal position to form a

stable hydrogen bond between each other but not so with the active site S or Se. It

therefore might not be a direct hydrogen bonding but the generally high density of

labile protons in the surrounding of UP or C

P, respectively, that enforces its

dissociation [27]

. It is further tempting to speculate that the surface-exposed

residues Gln and Trp are more important for the polarization of the peroxy bond

than for the activation of CP or U

P. Finally, a few exceptions from the canonical

triad / tetrad concept have been reported: The canonical Gln is replaced by Glu in

poplar GPx [53]

and by Ser in human GPx8 [23]

and could be replaced by Gly

without loss of activity in the GPx of Chinese cabbage [54]

, which reveals a certain

plasticity of the otherwise strictly conserved active site. Collectively, structural,

genetic and functional investigations have unraveled some important features of

(seleno) cysteine activation in GPx catalysis, but we are still far from a conclusive

concept. The extreme efficiencies of the magic sulphur and/or selenium atoms still

remain enigmatic.

1.2.4.3� Chemical Nature of Oxidized GPx

Another puzzle in GPx catalysis is the precise chemical nature of the oxidized


9

enzyme. In Eqs.(1.7) – (1.9) and (1.12), it is boldly shown as a sulfenic or

selenenic acid derivative of CP or U

P, respectively. In fact, this assumption is little

else but a postulate based on the stoichiometry of the reaction of one (seleno)

cysteine residue with one hydroperoxide molecule. Admittedly, the oxidation of

cysteine residues to sulfenic acids in proteins is not uncommon and has been

amply demonstrated to occur in the analogous peroxiredoxin catalysis [55]

. With

the selenoperoxidases, however, the situation is less clear. The postulated

selenenic acid form has so far never been demonstrated experimentally. In X-ray

crystallography the selenium of GPx1 was seen as seleninic acid [52]

. Instead, by

mass spectrometry oxidized GPx4 [20]

and GPx1 [56]

consistently showed a

molecular mass that was lower than that of the reduced enzyme by two mass units.

This finding would be compatible with elimination of H2O from the postulated

selenenic acid form and, in analogy to the catalysis of atypical 2-cysteine

peroxiredoxins, might be interpreted as indicating the formation of an

intramolecular selenyl-sulfide bond. There is, however, no cysteine residue in

GPx1 or GPx4 that could serve as such CR. Alternatively, an initially-formed

selenenic acid could react with a nearby amino, imino or amido group in analogy

to the redox cycle of the GPx mimic ebselen (2-phenyl-1,2-benzisoselenazol-

3(2H)-one) [57]

. An analogous sulfenamide bond has been identified in oxidized

protein tyrosine phosphatases such as PTP1B [58, 59]

and PTPα [60]

by X-ray

crystallography. Like the selenyl-amide bond on oxidized ebselen, the

sulfenamide in the PTPs is readily reduced by GSH [58-60]

, and it is therefore

tempting to speculate that the first intermediate of the GPx cycle is indeed a

selenyl-amide formed between the active site selenium and one of the triad/tetrad

components. However, none of the suspected Se-N bonds could so far be detected

by systematic mass spectrometry investigations, nor were they revealed by X-ray

studies. It therefore appears wise to address oxidized GPx as a “selenenic acid

equivalent” until its chemical nature has been clarified.

1.2.4.4� Structures and Substrate Specificities

As mentioned above, GPx1 is highly specific to GSH. Its GSH specificity has

been attributed to a lysine residue (K91´) and 4 arginine residues (R57, R103,

R184 and R185 in bovine GPx1) which surround the active site selenium and

serve to successively direct the two GSH molecules into an orientation that allows

reaction of the GSH sulphur with the selenium [61]

. This view has been corroborated

by modeling and molecular dynamics calculations [62]

but has so far not been

verified by mutagenesis studies (Fig. 1.2). Therefore, the relative importance of

the five basic residues must still be rated as uncertain. A contribution of these

residues to GSH binding is, however, also supported by the circumstantial

evidence that their deletion or replacement by non-equivalent residues, as is

observed in members of GPx subfamilies other than GPx1, leads to gradual or

complete loss of GSH specificity. The GPx2 subfamily has three of these residues


10

conserved and is presumed to be functionally closest to GPx1 (Fig. 1.2). In GPx3

only two of the arginines are conserved, which complies with a smaller k´+2

for

GSH [63]

and discrete reactivity with other substrates such as thioredoxin and

glutaredoxin [14, 64]

. None of the residues is conserved in GPx4 and indeed its

specificity is rather degenerated covering a variety of low molecular mass thiols as

well as protein thiols. However, also in GPx4 the active site selenium is flanked

by basic residues which may explain its relative preference for GSH [22, 27]

.

(a) (b) (c)

Fig. 1.2.� Structural basis for the GSH specificity of GPx1. The figure shows molecular models

of the ground state enzyme (a) with major residues implicated in GSH binding marked, the

catalytic intermediate formed according to Eq.(1.9), in which a glutathionyl residue (green;

residues given for GSH) is covalently bound to the active site selenium (b), and (c) the complex

of the intermediate with the second GSH molecule (magenta, GSH residues given for second

GSH), from which the ground state enzyme is regenerated according to Eq.(1.10). In essence, the

glycine carboxylic function of the first reacting GSH is assumed to bind to the guanidino group

of R57 and the cysteinyl carbonyl to R184, while the negative γ-glutamyl tail is attracted by R103

and by K91´ of an adjacent subunit (carbon atoms in brown). Thereby the sulphur of GSH is

oriented towards the enzyme’s selenium to form the intermediate (b). As demonstrated in (c), the

second GSH competes with the bound GSH for R184 with its glycine carboxyl, while its γ-

glutamyl residue is also attracted by the excess positive charge provided by R103 and K91´. R185

may facilitate product release. For description of methodology and further details see ref. [62]

The architecture typical of the GPx1 subfamily is not at all conserved in

GPx4 and neither in mammalian GPx6, GPx7 and GPx8 for which the donor

specificities have not yet been worked out. It is neither conserved in the non-

mammalian GPx homologues that are specific for redoxin-type proteins. A

consensus sequence that complies with thioredoxin specificity as well as the

peroxiredoxin-type mechanism has been delineated. It comprises an obligatory CR

located in a flexible loop [14]

. The mode of (thio)redoxin binding, however,

remains to be elucidated.

At least equally vague are the ideas on structural features that determine the

hydroperoxide specificity of GPxs. According to all available structures, the

selenium of UP or the sulphur of C

P, respectively, is surface-exposed and should

therefore be able to react with any kind of sterically accessible hydroperoxide. As

a rule, however, only monomeric GPxs such as GPx4 and the GPx-type thioredoxin

peroxidase of Drosophila melanogaster have been shown to reduce hydroperoxy

1.3� Biological Roles of Individual Glutathione Peroxidases

11

groups of complex lipids efficiently. In contrast, GPx1 does not at all reduce

hydroperoxides of phospholipids, nor any other hydroperoxy groups that are

integrated into biomembranes [65, 66]

, although otherwise its spectrum of hydroperoxide

substrates is very broad [67]

. GPx3 apparently adopts an intermediate position in

reducing phosphatidylcholine hydroperoxide at a comparatively low rate [45]

(for a

recent compilation of rate constants see ref. [27]). The failure of tetrameric GPxs

to interact efficiently with hydroperoxides of lipids in micelles or membranes

might be due to sterical hindrance, since their active sites are located in a flat

valley built by the subunit interfaces, while in monomeric ones it forms an easily

accessible flat surface. Alternatively, the clear preference of GPx4 for

phospholipid hydroperoxides has been attributed to an extended area of positive

surface charges which might attract negatively-charged lipids [27]

.


The discrete differences in substrate specificities, subcellular localization and

tissue distribution of the GPxs are reflected in distinct biological roles which, in

part at least, have been unraveled by genetic techniques.

1.3.1� GPx1

GPx1 is the most abundant GPx in mammals. It appears to be predominantly

localized in the cytosol and the matrix space of mitochondria [68]

. Since its

discovery, GPx1 has been considered as an enzyme that protects the organism

against oxidative damage [69]

. Impaired GPx1 function was considered to be the

ultimate pathogenic consequence in genetic disorders of the pentose-phosphate

shunt that result in insufficient GSH regeneration and, in consequence, lead to

hemolytic disorders [67, 70]

. Furthermore, a related phenomenon, H2O

2-induced

hemolysis in selenium-deficient rats, provided the key to discover the selenoprotein

nature of GPx1 [2, 3]

. The view that GPx1 is primarily an emergency device to

counteract a hydroperoxide challenge was corroborated by inverse genetics. GPx1

knock-out mice develop normally, even grow faster and tolerate elevated oxygen

tension [71]

. This lack of any overt phenotype reminds one of human genetic

deficiencies in GSH regeneration which were rated as non-diseases [70]

. “Patients”

with glucose-phosphate dehydrogenase deficiency for instance are asymptomatic

if not challenged by pro-oxidant drugs or other xenobiotics. Similarly, the genetic

defect of GPx1−/−

mice only becomes evident upon an oxidative challenge such as

exposure to redox-cycling herbicides and to lipopolysaccharides that trigger an

oxidative burst in phagocytes [30]

. However, the GPx1−/−

mouse also demonstrated

that the seemingly-normal life without GPx1 is threatened by viral infections. The


12

GPx1−/−

mouse proved to be a convincing model for the classical human selenium-

deficiency syndrome Keshan disease. The GPx1−/−

mice, like selenium-deficient

mice, when exposed to a non-virulent Coxsackie strain, developed a cardio-

myopathy that resembled Keshan disease, and it was shown that the virus had

mutated into a virulent form [72, 73]

. Thus, impaired peroxide metabolism in genetic

GPx1 deficiency and selenium deficiency triggers the same events: an increased

mutation rate and persistence of the more aggressive mutant viral strains (The

similarity of knock-out of the GPx1 gene and alimentary selenium deficiency

is best explained by the fact that GPx1 ranks low in the hierarchy of the

selenoproteins [29, 30]

).

A protection against viral infection by GPx1 should not be generally expected.

Selenium-deficient and, thus, GPx1-deficient mice survived an influenza virus

infection significantly better than selenium-adequate ones due to an enhanced

immune response [74]

. In fact, an increasing number of odd observations indicate

that GPx1 is not a benign enzyme under all circumstances: Overexpression of

GPx1

i) reduced TNFα-induced NFκB activation [75]

;

ii) enhanced viral spread in HIV-infected tissue culture [76]

;

iii) promoted acetaminophen toxicity [77]

;

iv) increased DMBA/TPA-induced carcinogenesis [78]

and

v) triggered obesity and insulin resistance in transgenic animals [79]

.

Inversely, disruption of the GPx1 gene rendered hepatocytes resistant to

peroxynitrite [80]

, although the enzyme has been shown to reduce peroxynitrite [81]

,

and protected mice against kainic-acid-induced seizures [82]

. Collectively, these

poorly-explained findings suggest that the concept of fighting oxidative stress is a

simplification even for GPx1.

Regulation of GPx1 expression does not fully comply with a merely antioxidant

role of GPx1 either. The gene contains functional oxygen-responsive and tumor-

promoting-agent-responsive elements (ORE 1 and TRE, respectively) [83, 84]

, the

latter being activated by AP-1 [83]

. As has long been known [85]

, however, GPx1 is

also up-regulated by estrogens, which is probably mediated by NFκB [86]

, and

appears to be under the control of diverse transcription factors such as PU1 [87]

,

RXR [88]

and p53 [89, 90]

.

1.3.2� GPx2

GPx2 has so far not been extensively characterized as an isolated enzyme, but

presumably displays a substrate specificity spectrum that appears largely identical to

that of GPx1 [91]

. It is primarily found in the epithelial lining of the gastro-intestinal

tract and was originally claimed to protect against food-born hydroperoxides [91]

. In

humans, in contrast to rodents, it is also found in the liver [91]

. Beyond this, it has

been detected in several transformed epithelial cells [30, 91]

and was found to be up-


13

regulated in certain cancer tissues [92]

. The tissue distribution within the intestine is

quite unique. It is seen in Paneth cells [93]

, particularly enriched in the crypt

epithelium and gradually declines towards the tip of the villi [93, 94]

. Within the

epithelial cell it partially co-localizes with markers of the Golgi system [93]

.

Knock-out of the GPx2 gene did not yield any obvious phenotype. However, a

double knock-out of GPx2 and GPx1 results in an early on-set ileo-colitis, which

resembled human Crohn’s disease [95]

, and after 6 – 9 months in inflammation-

driven manifestation of malignant tumors [96]

. The observation suggests a mutual

complementation of the two isoforms, which is in line with their identical

specificity, while the peculiar role of GPx2 has to be attributed to its tissue

distribution. In the crypt ground it appears to guarantee cellular protection and

continuous proliferation, whereas towards the villi, which are poor in both

GPx1and GPx2, physiological apoptosis is enabled. Therefore, the concerted

action of GPx1 and GPx2 might be pivotal to the delicate balance of proliferation

and apoptosis and thus to the continuous regeneration of the gut epithelium.

This concept is largely supported by studies on the transcriptional regulation

of GPx2. GPx2 is a direct target of the transcription factor p63, which induces cell

cycle arrest and inhibits apoptosis [97]

. Its transcription is further regulated by

retinoic acid [98]

, the transcription factors Nrf2 [99]

and β-catenin [100]

. Wnt-

signaling-dependent activation of β-catenin associates GPx2 with proliferative

potential, while Keap-1-mediated Nrf2-activation clearly places GPx2 in the

context of protective and anti-carcinogenic proteins.

1.3.3� GPx3

GPx3 is a typical extracellular glycosylated protein [101]

with a largely unknown

function. The primary site of biosynthesis is the proximal tubulus of the kidney [102]

,

from where it is secreted into the circulation, but it is also expressed in adipose

tissue [103]

and in lung epithelial cells [104]

, from where it reaches the lining fluid [105]

.

Its hydroperoxide specificity resembles that of GPx1 [63]

, although it also reduces

complex lipid hydroperoxides at a low rate [45]

.

GPx3 is generally considered not only as an “antioxidant” enzyme, but also as

an “orphan” enzyme, since it has to work in an environment that is very low in the

reducing substrate GSH and devoid of any known GSH-regenerating system. Its

capacity to fight an extracellular oxidative stress would indeed be exhausted after

a few catalytic cycles. Its biological role thus remains elusive. Some putative roles

are being discussed.

i) The low capacity of the extracellular GPx system may suffice to prevent

LDL oxidation and thus atherogenesis; but the substrate specificity of GPx3 is not

ideal for meeting this task.

ii) GPx3 might down-regulate the activity lipoxygenases (LOX) and cyclo-

oxygenases (COX) by lowering the extracellular peroxide tone, thereby preventing


14

undue formation of leukotrienes and prostaglandins and, in consequence,

inflammatory responses upon irrelevant stimuli.

iii) Finally, oxidized GPx3 itself might be a signaling molecule that “senses” a

local peroxide release and, by means of its oxidized selenium, alarms remote

redox-sensitive receptors [30]

.

Unfortunately, these ideas remain speculative, since GPx3 knock-out mice

proved to be asymptomatic and did not so far respond to any challenge (including

lipopolysaccharide challenge) in any unusual way [106]

.

Analysis of transcriptional activation of the GPx3 gene does not reveal any

consistent picture either: the redox-sensitive transcription factor AP-1 [104]

, SP-1

and the hypoxia-induced factor HIF-1 [107]

, as well as the corticoid-responsive

factor C/EBPδ in adipocytes [103]

, have been implicated in the regulation of GPx3

expression, and the gene further contains an anti-oxidant-responsive element

(ARE) and a heavy-metal-responsive element (MRE) [107]

.

1.3.4� GPx4

GPx4 was first described as a peroxidation inhibiting protein (PIP) that protected

phosphatidylcholine liposomes and biomembranes from peroxidatic degradation in

the presence of glutathione, then shown to be a monomeric GPx with an unusually

broad specificity spectrum and designated as phospholipid hydroperoxide

glutathione peroxidase (PHGPx) [108, 109]

. Its function has for long been seen in the

protection of biomembranes against oxidative degradation. Over the last two

decades, however, a bewildering number of additional, in part fully-unexpected

functions were reported, making GPx4 a chameleon of this protein family.

First of all, the GPx4 gene is translated into three different forms: a cytosolic

(cGPx4), a mitochondrial (mGPx4) and a nuclear one (nGPx4). As mature proteins

cGPx4 and mGPx4 are identical but mGPx4, by use of alternate upstream

transcription initiation and translation sites, is synthesized with a mitochondrial

leader sequence which is cleaved during uptake into mitochondria [110]

. The use of

an alternate downstream promoter and translation start that is located in the first

intron of the gene results in nGPx4 [111-114]

, which has a distinct N-terminus that

comprises the nuclear leader sequence [115]

. All three forms are found in the entire

organism, but most abundantly in testis.

Despite substantial efforts, the regulation of tissue- and stage-specific biosynthesis

of the GPx4 forms is far from clear. The upstream promoter region contains binding

sites for NF-Y [116]

, C/EBPε [117]

, GATA, SP1, AP2 and many putative binding sites

for unknown factors [114]

, while in the nGPx promoter region of the first intron

binding sites for STRE, SP1, CREB, AP1 [114]

and CREM-τ [113]

could be identified.

The multiple transcriptional regulators differentially affect basic expression of the

GPx4 isoforms in a complex manner [114]

and may be further complicated by

translational regulation [118]

. Regulated expression of GPx4 has not been systematically


15

analyzed. GPx4 was found down-regulated in polymorphonuclear leukocytes

immediately after activation, but up-regulated later, suggesting a response to oxidative

stress [119]

. Similarly, GPx4 is up-regulated by pro-inflammatopry cytokines

such as TNFα in PMNs [120]

, HL60 cells [117]

and interleukin-1β in umbilical

vein endothelial cells [121]

. In the male reproductive system GPx4 is negatively

affected by estradiol [122]

, but estrogen up-regulates GPx4 in the oviduct [123]

. GPx4

in testis is under indirect control of gonadotropin [124]

or testosterone [125]

,

respectively. Considering this confusing scenario and the many unknowns, it

cannot be a surprise that the unusual tissue- and stage-specific expression pattern

of GPx4 isoforms is still poorly understood.

More insight into the physiological role of GPx4 is provided by inverse

genetics. Total disruption of the GPx4 gene resulted in embryonic lethality

between day 7.5 and 8.5 p.c. [126]

, revealing a pivotal role in embryonic

development of this GPx in contrast to all others investigated so far. Since a

specific knock-out of nGPx [127]

and mGPx [128]

yielded viable off-springs, the

developmental role has to be attributed to the cytosolic form. The reasons for the

unique essentiality of cGPx4 are not entirely clear. An intriguing hypothesis

attributes the phenomenon to the anti-apoptotic potential of the enzyme [129]

:

12/15-lipoxygenase is an enzyme that has for long been implicated in oxidative

destruction of biomembranes during cellular differentiation [130]

and triggers

apoptosis via a particular pathway mediated by the apoptosis inducing factor

AIF [129]

. The 12/15-lipoxygenase peroxidizes lipids within membranes and thus

creates the preferred substrates of GPx4, which are not efficiently reduced by any

of the other GPxs. Accordingly, neuron-specific GPx4 depletion caused neuro-

degeneration in vivo and ex vivo, which was antagonized by AIF silencing or

inhibition of 12/15-lipoxygenase [129]

. The antagonistic couple GPx4/12/15-

lipoxygenase thus sustains a delicate redox balance that appears to be of

outstanding importance for apoptosis-mediated tissue remodelling, as is required

in embryogenesis. In contrast to total GPx4 knock-out, specific deletion of nGpx4

proved to be compatible with normal development. However, in line with the

presumed role of nGPx in chromatin condensation [115]

, spermatozoa of nGPx−/−

mice showed defective chromatin condensation and morphologically abnormal

heads. These anomalies, however, were not associated with impaired fertility, nor

with any other obvious phenotype [127]

. In contrast, specific knock-out of the

mitochondrial GPx4 resulted in abrogation of male fertility in otherwise normal

animals [128]

. Therefore, it is clearly the mGPx4 which “moonlights for fertility”

during spermiogenesis.

The relevance of GPx4 to male fertility has been amply reviewed [33, 34]

.

However, some related aspects shall here be recalled, because they opened up a

new chapter of GPx research that might be relevant to various enigmatic findings.

The key finding that linked selenium research to andrology was the moonlighting

nature of GPx4. The enzyme is abundantly synthesized in early spermatids and is

there detectable as an active GPx, whereas in mature spermatozoa GPx4 is

enzymatically inactive and presents itself as a structural protein that makes up

more than 50% of the mitochondrial capsule in the mid piece [21]

. GPx4 deficiency,


16

due to whatever reason, causes instability of the mid piece architecture leading to

impaired sperm mobility, morphological alterations up to loss of sperm tail and, in

consequence, lack of fertilisation capacity [33, 34, 131, 132]

. Mechanistically, the switch

from an active peroxidase to an almost insoluble structural protein “moonlighting”

is triggered by a loss of GSH in late spermiogenesis, which results from a still

poorly understood burst of hydroperoxide generation followed by oxidation of

GSH, export of GSSG and extracellular degradation by γ-GT. Due to the shortage

of GSH, GPx4 now reacts with a particular SH group of itself, which results in

linear GPx polymers [22]

, uses cysteine-rich proteins, in particular those with PCCP

motifs as alternate substrates [19]

, and ultimately integrates itself into a three-

dimensional polymerisation product that is cross-linked by selenyl-sulfide and

disulfide bonds and has the physical consistency of hair or finger nails. Thus, the

unusual moonlighting phenomenon does not require any other chemistry than that

describing the basic steps of GPx catalysis (Eqs.(1.7) – (1.10), above). It may,

however, be stressed that the enzyme also changes its biological destiny from a

peroxide-eliminating peroxidase to a thiol-oxidizing protein that just makes use of

hydroperoxides to build up a biological structure that is indispensable for an

appropriate sperm function. Also, GPx4, which like its congeners is an anti-

apoptotic protein [128, 129, 133, 135]

, here becomes pro-apoptotic in creating the mature

spermatozoon which is an almost completely oxidized cell prone to die within a

couple of days. Moreover, the process for the first time revealed that a GPx can

indeed act as a protein thiol-modifying agent, a phenomenon that has meanwhile

become known as an important regulatory principle involving cysteine GPx

homologues and analogous peroxiredoxins [32, 136]

.

Another important role of GPx4 is the suppression of inflammatory responses

via different mechanisms. In vitro, it shares with other GPxs the ability to silence

lipoxygenases and cyclo-oxygenases. All these enzymes have to become oxidized

by an ROOH to become active, as first described for ROOH removal by GPx1 [137]

.

In vivo, however, GPx4 appears to be the most efficient suppressor of lipoxygenase

activity, probably because the favourite lipophilic GPx4 substrates are better LOX

and COX activators than H2O

2

[138, 139]

. Beyond this, GPx4 was reported to

suppress the expression of pro-inflammatory proteins such as COX-2 [140]

and

VCAM-1 [141]

. Finally, a small overexpression of GPx4 inhibited UV-light-induced

NFκB activation in skin fibroblasts [142]

and completely suppressed interleukin-1-

induced activation of NFκB in arterial smooth muscle cells, an effect that could

not be mimicked by large variations of GPx1 activity [143]

, the peculiar specificity

of GPx4 remaining unexplained.

In short, its broad substrate specificity allows GPx4 to adopt multiple, in part

contrasting, biological roles, which depend on the metabolic context.

i) As a broad spectrum glutathione peroxidase it prevents oxidative damage

and apoptosis, in particular if the latter is due to enzymatic lipid peroxidation.

Thereby it is likely essential for embryogenesis.

ii) As a “protein thiol peroxidase”, it is rather pro-apoptotic.

iii) By means of protein thiol peroxidase activity, its nuclear form contributes

to chromatin condensation.

1.4� Conclusions and Perspectives

17

iv) During spermiogenesis, the protein thiol peroxidase activity is believed to

be essential for the oxidation of cysteine-rich proteins and is definitely pivotal to

its self-transformation into a structural protein and thus to sperm function.

v) It is clearly an anti-inflammatory protein by silencing the key enzymes for

the production of pro-inflammatory lipid mediators, by suppressing the expression

of pro-inflammatory enzymes and cytokines and by interfering pro-inflammatory

signalling cascades.

1.3.5� GPx5–GPx8

GPx5 is a non-selenium extracellular GPx that is found in the epididymal fluid [144]

.

It is synthesized under androgen control possibly mediated by the polyoma

enhancer activator protein PEA3 [145]

. GPx5−/−

mice did not display any obvious

phenotype, remained fertile, but showed increased miscarriages and malformations

in their off-springs, suggesting protection against DNA damage by GPx5 [146]

.

GPx6 was first described as a putative odorant-binding or metabolizing protein [147]

and much later identified as selenium-containing GPx in man, while it is a non-

selenium GPx in rodents [148]

. Its function is unknown. GPx7 is a monomeric non-

selenium GPx that reportedly has low GPx activity but is nevertheless implicated

in the management of oxidative stress [149]

. GPx8 finally is a very remote relative

of the family of which little else than the sequence is known (UniProt/Swiss-Prot:

Q8TED1).

1.4� Conclusions and Perspectives

As outlined above, GPx catalysis has amply demonstrated its versatile utility and

keeps the promise of further surprises, since not even the biological roles of the

mammalian GPxs have been fully elucidated. In fact, for most of them the

biological relevance remains elusive or is based on circumstantial evidence such

as factors regulating their transcription, in vitro activities or phenomena resulting

from gene disruption or overexpression. Collectively, GPx1 appears to be the

enzyme in charge of preventing damage from an H2O

2 challenge, but the

pathology induced by overexpression of this antioxidant enzyme reveals its

involvement in metabolic regulation. For sure, the other GPxs should not be

simply considered to back-up GPx1. They rather deserve interest as potential

redox sensors or regulators or, like GPx4, find their destiny in highly specialized

cellular differentiation processes.

To analyze how this bewildering scope of biological effects is brought about

by a single type of catalysis, will remain a challenge for quite a while. Some

mechanistic principles have been worked out.


18

i) Draining the flux of H2O

2 that is, for example, required for insulin signaling,

might physiologically guarantee a balanced hormone response, but must result in

dysregulation, if it happens in excess. This most simple form of metabolic

interference results from the basic catalytic potential of every GPx to speed up the

reaction according to Eq.(1.1) which steadily competes with the oxidation of

susceptible SH groups in target proteins of signaling cascades.

ii) Such oxidative protein modification will first result in sulfenic acid

formation in analogy to Eq.(1.7) and will be followed by internal disulfide

formation, if possible, by glutathionylation or by sulfenamide formation, as

exemplified in phosphatases, associated with positive or negative impact on the

activity of target proteins.

iii) An oxidized GPx can by itself modify a protein SH group, as demonstrated

for mammalian GPx4 or, in the context of transcriptional regulation, for the GPx-

type Orp-1 protein of yeast.

iv) Finally, the products of ROOH reduction such as HETEs, B-type

leukotrienes, prostanoids and other final lipoxygenase products are by no means

inert compounds.

Which of these principles becomes physiologically relevant depends on the

specificity of the GPx involved, its cellular or subcellular distribution, steady state

concentrations of substrates and, most importantly, from the rate constants of the

regular GPx reaction in relation to those of the competing ones, which are largely

unknown. The analysis is further complicated by the co-existence of eight distinct

peroxiredoxins, which in part at least share activity and specificity with the GPx

family and are equally implicated in redox regulation [150]

. Finally, GPx- or

peroxiredoxin-mediated signals have to be reversed, if they are to build up

physiologically meaningful regulatory circuits, an aspect that has so far only been

addressed in a few cases [32, 136]

.

In summary, the realm of peroxidases is not likely an assembly of redundant

devices to shield us from the dangers of aerobic life, but a highly-specialized

system of enzymes that complement each other in response to changed fluxes of

particular peroxides due to exogenous stimuli or endogenous needs.

References

[1] Flohé L, Günzler WA, Schock HH (1973) FEBS Lett 32: 132

[2] Rotruck JT, Pope AL, Ganther HE, Swanson AB, Hafeman DG, Hoekstra WG

(1973) Science 179: 588

[3] Flohé L (2009) Biochim Biophys Acta 1790: 1389

[4] Tujebajeva RM, Copeland PR, Xu XM, Carlson BA, Harney JW, Driscoll

DM, Hatfield DL, Berry MJ (2000) EMBO Rep 1: 158

[5] Yoshizawa S, Böck A (2009) Biochim Biophys Acta 1790: 1404

[6] Xu XM, Carlson BA, Zhang Y, Mix H, Kryukov GV, Glass RS, Berry MJ,

References

19

Gladyshev VN, Hatfield DL (2007) Biol Trace Elem Res 119: 234

[7] Allmang C, Wurth L, Krol A (2009) Biochim Biophys Acta 1790: 1415

[8] Squires JE, Berry MJ (2008) IUBMB Life 60: 232

[9] Maiorino M, Roche C, Kiess M, Koenig K, Gawlik D, Matthes M, Naldini E,

Pierce R, Flohé L (1996) Eur J Biochem 238: 838

[10] Salinas G, Lobanov AV, Gladyshev VN (2006) In: Hatfield DL, Berry MJ,

Gladyshev VN (eds) Selenium, Its Molecular Biology and Role in Human

Health. 2nd Ed., Springer, New York: 355

[11] Bae YA, Cai GB, Kim SH, Zo YG, Kong Y (2009) BMC Evol Biol 9: 72

[12] Cossio-Bayugar R, Niranda E, Holman PJ (2005) Insect Biochem, Mol, Biol

35: 1378

[13] Gladyshev VN (2006) In: Hatfield DL, Berry MJ, Gladyshev VN (eds)

Selenium, Its Molecular Biology and Role in Human Health. 2nd Ed.,

Springer, New York: 99

[14] Maiorino M, Ursini F, Bosello V, Toppo S, Tosatto SC, Mauri P, Becker K,

Roveri A, Bulato C, Benazzi L, De Palma A, Flohé L (2007) J Mol Biol 365:

1033

[15] Schlecker T, Comini MA, Melchers J, Ruppert T, Krauth-Siegel RL (2007)

Biochem J 405: 445

[16] Delaunay A, Pflieger D, Barrault MB, Vinh J, Toledano MB (2002) Cell 111:

471

[17] Flohé L, Günzler W, Jung G, Schaich E, Schneider F (1971) Hoppe Seylers Z

Physiol Chem 352: 159

[18] Godeas C, Tramer F, Micali F, Roveri A, Maiorino M, Nisii C, Sandri G,

Panfili E (1996) Biochem Mol Med 59: 118

[19] Maiorino M, Roveri A, Benazzi L, Bosello V, Mauri P, Toppo S, Tosatto SC,

Ursini F (2005) J Biol Chem 280: 38395

[20] Tosatto SC, Bosello V, Fogolari F, Mauri P, Roveri A, Toppo S, Flohé L,

Ursini F, Maiorino M (2008) Antioxid Redox Signal 10: 1515

[21] Ursini F, Heim S, Kiess M, Maiorino M, Roveri A, Wissing J, Flohé L (1999)

Science 285: 1393

[22] Mauri P, Benazzi L, Flohé L, Maiorino M, Pietta PG, Pilawa S, Roveri A,

Ursini F (2003) Biol Chem 384: 575

[23] Toppo S, Vanin S, Bosello V, Tosatto SC (2008) Antioxid Redox Signal 10:

1501

[24] Lawrence RA, Burk RF (1976) Biochem Biophys Res Commun 71: 952

[25] Saito Y, Sato N, Hirashima M, Takebe G, Nagasawa S, Takahashi K (2004)

Biochem J 381: 841

[26] Schremmer B, Manevich Y, Feinstein SI, Fisher AB (2007) In: Flohé L,

Harris JR (eds) Peroxiredoxin Systems, Structure and Functions. Springer,

New York, Subcell Biochem, 44: 317

[27] Toppo S, Flohé L, Ursini F, Vanin S, Maiorino M (2009) Biochim Biophys

Acta 1790: 1486

[28] Trujillo M, Ferrer-Sueta G, Thomson L, Flohé L, Radi R (2007) In: Flohé L,

Harris JR (eds) Peroxiredoxin Systems, Structure and Functions. Springer,


20

New York, Subcell Biochem, 44: 83

[29] Sunde RA (2006) In: Hatfield, DL, Berry, MJ, Gladyshev VN (eds) Selenium,

Its Molecular Biology and Role in Human Health. 2nd Ed., Springer, New

York: 149

[30] Flohé L, Brigelius-Flohé R (2006) In: Hatfield DL, Berry MJ, Gladyshev VN

(eds) Selenium, Its Molecular Biology and Role in Human Health. 2nd Ed.,


[31] Brigelius-Flohé R (2006) Biol Chem 387: 1329

[32] Fourquet S, Huang ME, D'Autreaux B, Toledano MB (2008) Antioxid Redox

Signal 10: 1565

[33] Flohé L (2007) Biol, Chem 388: 987

[34] Boitani C, Puglisi R (2008) In: Cheng CY (ed) Molecular Mechanisms in

Spermatogenesis. Springer, New York: 65

[35] Nakagawa Y (2004) Biol Pharm Bull 27: 956

[36] Loscalzo J (2008) Cell Metab 8: 182

[37] Beck MA (2006) In: Hatfield DL, Berry MJ, Gladyshev VN (eds) Selenium,


York: 287

[38] Schmutzler C, Mentrup B, Schomburg L, Hoang-Vu C, Herzog V, Köhrle J

(2007) Biol Chem 388: 1053

[39] Schweizer U,Schomburg L (2006) In: Hatfield DL, Berry MJ, Gladyshev VN

(eds) Selenium, Its Molecular Biology and Role in Human Health. 2nd Ed.,


[40] Hatfield DL, Berry MJ, Gladyshev VN (eds) (2006) Selenium, Its Molecular

Biology and Role in Human Health. 2nd Ed., Springer, New York

[41] Flohé L, Harris JR (2007) In: Flohé L, Harris JR (eds) Peroxiredoxin Systems,

Structure and Functions. Springer, New York, Subcell Biochem, 44: 1

[42] Ostwald W (1906) Leitlinien der Chemie, Sieben gemeinverständliche,

Vorträge aus der Geschichte der Chemie. Akademische Verlagsgesellschaft

m, b H, Leipzig: 274

[43] Dalziel K (1957) Acta Chem Scand 11: 1706

[44] Flohé L, Loschen G, Günzler WA, Eichele E (1972) Hoppe Seylers Z Physiol

Chem 353: 987

[45] Takebe G, Yarimizu J, Saito Y, Hayashi T, Nakamura H, Yodoi J, Nagasawa

S, Takahashi K (2002) J Biol Chem 277: 41254

[46] Maiorino M, Roveri A, Gregolin C, Ursini F (1986) Arch Biochem Biophys

251: 600

[47] Kehr S, Malinouski M, Finney L, Vogt S, Labunskyy VM, Kasaikina MV,

Carlson BA, Zhou Y, Hatfield DL, Gladyshev VN (2009) J Mol Biol 389:

808

[48] Winterbourn CC, Metodiewa D (1999) Free Radic Biol Med 27: 322

[49] Maiorino M, Aumann KD, Brigelius-Flohé R, Doria D, van den Heuvel J,

McCarthy J, Roveri A, Ursini F, Flohé L (1995) Biol Chem Hoppe Seyler

376: 651

[50] Ma LH, Takanishi CL, Wood MJ (2007) J Biol Chem 282: 31429

References

21

[51] Rocher C, Lalanne JL, Chaudière J (1992) Eur J Biochem 205: 955

[52] Epp O, Ladenstein R, Wendel A (1983) Eur J Biochem 133: 51

[53] Koh CS, Didierjean C, Navrot N, Panjikar S, Mulliert G, Rouhier N, Jacquot

JP, Aubry A, Shawkataly O, Corbier C (2007) J Mol Biol 370: 512

[54] Jung BG, Lee KO, Lee SS, Chi YH, Jang HH, Kang SS, Lee K, Lim D,

Yoon SC, Yun DJ, Inoue Y, Cho MJ, Lee SY (2002) J Biol Chem 277:

12572

[55] Poole L (2007) In: Flohé L, Harris JR (eds) Peroxiredoxin Systems, Structure

and Functions. Springer, New York, Subcell Biochem, 44: 61

[56] Ursini F (2009) Personal communication

[57] Sies H (1994) Methods Enzymol 234: 476

[58] Salmeen A, Andersen JN, Myers MP, Meng TC, Hinks J A, Tonks NK,

Barford D (2003) Nature 423: 769

[59] van Montfort RL, Congreve M, Tisi D, Carr R, Jhoti H (2003) Nature 423:

773

[60] Yang J, Groen A, Lemeer S, Jans A, Slijper M, Roe SM, den Hertog J,

Barford D (2007) Biochemistry 46: 709

[61] Ursini F, Maiorino M, Brigelius-Flohé R, Aumann KD, Roveri A,

Schomburg D, Flohé L (1995) Methods Enzymol 252: 38

[62] Aumann KD, Bedorf N, Brigelius-Flohé R, Schomburg D, Flohé L (1997)

Biomed Environ Sci 10: 136

[63] Esworthy RS, Chu FF, Geiger P, Girotti AW, Doroshow JH (1993) Arch

Biochem Biophys 307: 29

[64] Bjørnstedt M, Xue J, Huang W, Akesson B, Holmgren A (1994) J Biol Chem

269: 29382

[65] Grossmann A, Wendel A (1983) Eur J Biochem 135: 549

[66] Sevanian A, Muakkassah-Kelly SF, Montestruque S(1983) Arch Biochem

Biophys 223: 441

[67] Flohé L (1989) In: Dolphin D, Poulson R, Avramovic O (eds) Glutathione:

Chemical,Biochemical and Medical Aspects - Part A. John Wiley & Sons,

Inc: 643

[68] Flohé L, Schlegel W (1971) Hoppe Seylers Z Physiol Chem 352: 1401

[69] Mills GC (1957) J biol Chem 229: 189

[70] Beutler E (1983) Biomed Biochim Acta 42: S234

[71] Ho YS, Magnenat JL, Bronson RT, Cao J, Gargano M, Sugawara M, Funk

CD (1997) J Biol Chem 272: 16644

[72] Beck MA, Shi Q, Morris VC, Levander OA (1995) Nat Med 1: 433

[73] Beck MA (2006) In: Hatfield DL, Berry MJ, Gladyshev VN (eds) Selenium,


York: 287

[74] Li W, Beck MA (2007) Exp Biol Med (Maywood) 232: 412

[75] Kretz-Remy C, Mehlen P, Mirault ME, Arrigo AP (1996) J Cell Biol 133:

1083

[76] Sandstrom PA, Murray J, Folks TM, Diamond AM (1998) Free Radic Biol

Med 24: 1485


22

[77] Mirochnitchenko O, Weisbrot-Lefkowitz M, Reuhl K, Chen L, Yang C,

Inouye M (1999) J Biol Chem 274: 10349

[78] Lu YP, Lou YR, Yen P, Newmark HL, Mirochnitchenko OI, Inouye M,

Huang MT (1997) Cancer Res 57: 1468

[79] McClung JP, Roneker CA, Mu W, Lisk DJ, Langlais P, Liu F, Lei XG, (2004)

Proc Natl Acad Sci USA 101: 8852

[80] Fu Y, Sies H, Lei XG (2001) J Biol Chem 276: 43004

[81] Briviba K, Kissner R, Koppenol WH, Sies H (1998) Chem Res Toxicol 11:

1398

[82] Jiang D, Akopian G, Ho YS, Walsh JP, Andersen JK (2000) Exp Neurol 164:

257

[83] Jornot L, Junod AF (1997) Biochem J 326: 117

[84] Merante F, Altamentova SM, Mickle DA, Weisel RD, Thatcher BJ, Martin

BM, Marshall JG, Tumiati LC, Cowan DB, Li RK (2002) Mol Cell Biochem

229: 73

[85] Pinto RE, Bartley W (1969) Biochem J 115: 449

[86] Vina J, Sastre J, Pallardo FV, Gambini J, Borras C (2008) Biol Chem 389:

273

[87] Throm SL, Klemsz MJ (2003) J Leukoc Biol 74: 111

[88] Wu Y, Zhang X, Bardag-Gorce F, Robel RC, Aguilo J, Chen L, Zeng Y,

Hwang, K, French SW, Lu SC, Wan YJ (2004) Mol Pharmacol 65: 550

[89] Gladyshev VN, Factor VM, Housseau F, Hatfield DL (1998) Biochem

Biophys Res Commun 251: 488

[90] Tan M, Li S, Swaroop M, Guan K, Oberley LW, Sun Y (1999) J Biol Chem

274: 12061

[91] Chu FF, Doroshow JH, Esworthy RS (1993) J Biol Chem 268: 2571

[92] Brigelius-Flohé R, Kipp A (2009) Biochim Biophys Acta 1790: 1555

[93] Florian S, Wingler K, Schmehl K, Jacobasch G, Kreuzer OJ, Meyerhof W,

Brigelius-Flohé R (2001) Free Radic Res 35: 655

[94] Chu FF, Esworthy RS (1995) Arch Biochem Biophys 323: 288

[95] Esworthy RS, Aranda R, Martin MG, Doroshow JH, Binder SW, Chu FF

(2001) Am J Physiol Gastrointest Liver Physiol 281: G848

[96] Lee DH, Esworthy RS, Chu C, Pfeifer GP, Chu FF (2006) Cancer Res 66:

9845

[97] Yan W, Chen X (2006) J Biol Chem 281: 7856

[98] Chu FF, Esworthy RS, Lee L, Wilczynski S (1999) J Nutr 129: 1846

[99] Banning A, Deubel S, Kluth D, Zhou Z, Brigelius-Flohé R (2005) Mol Cell

Biol 25: 4914

[100] Kipp A, Banning A, Brigelius-Flohé R (2007) Biol Chem 388: 1027

[101] Takahashi K, Avissar N, Whitin J, Cohen H (1987) Arch Biochem Biophys

256: 677

[102] Avissar N, Ornt DB, Yagil Y, Horowitz S, Watkins RH, Kerl EA,

Takahashi K, Palmer IS, Cohen HJ (1994) Am J Physiol 266: C367

[103] Yamasaki T, Tahara K, Takano S, Inoue-Murayama M, Rose MT,

Minashima T, Aso H, Ito S (2006) Cell Tissue Res 326: 139

References

23

[104] Comhair SA, Bhathena PR, Farver C, Thunnissen FB, Erzurum SC (2001)

Faseb J 15: 70

[105] Avissar N, Finkelstein JN, Horowitz S, Willey JC, Coy E, Frampton MW,

Watkins RH, Khullar P, Xu YL, Cohen HJ (1996) Am J Physiol 270: L173

[106] Burk RF (2009) Personal communication

[107] Bierl C, Voetsch B, Jin RC, Handy DE, Loscalzo J (2004) J Biol Chem 279:

26839

[108] Ursini F, Maiorino M, Valente M, Ferri L, Gregolin C (1982) Biochim

Biophys Acta 710: 197

[109] Schuckelt R, Brigelius-Flohé R, Maiorino M, Roveri A, Reumkens J,

Strassburger W, Ursini F, Wolf B, Flohé L (1991) Free Radic Res Commun

14: 343

[110] Arai M, Imai H, Sumi D, Imanaka T, Takano T, Chiba N, Nakagawa Y

(1996) Biochem Biophys Res Commun 227: 433

[111] Maiorino M, Scapin M, Ursini F, Biasolo M, Bosello V, Flohé L (2003) J

Biol Chem 278: 34286

[112] Moreno SG, Laux G, Brielmeier M, Bornkamm GW, Conrad M (2003) Biol

Chem 384: 635

[113] Tramer F, Vetere A, Martinelli M, Paroni F, Marsich E, Boitani C, Sandri G,

Panfili E (2004) Biochem J 383: 179

[114] Imai H, Saito M, Kirai N, Hasegawa J, Konishi K, Hattori H, Nishimura M,

Naito S, Nakagawa Y (2006) J Biochem 140: 573

[115] Pfeifer H, Conrad M, Roethlein D, Kyriakopoulos A, Brielmeier M,

Bornkamm GW, Behne D (2001) Faseb J 15: 1236

[116] Huang HS, Chen CJ, Chang WC (1999) FEBS Lett 455: 111

[117] Hattori H, Imai H, Kirai N, Furuhama K, Sato O, Konishi K, Nakagawa Y

(2007) Biochem J 408: 277

[118] Ufer C, Wang CC, Fahling M, Schiebel H, Thiele BJ, Billett EE, Kuhn H,

Borchert A (2008) Genes Dev 22: 1838

[119] Hattori H, Imai H, Hanamoto A, Furuhama K, Nakagawa Y (2005)

Biochem J 389: 279

[120] Hattori H, Imai H, Furuhama K, Sato O, Nakagawa Y (2005) Biochem

Biophys Res Commun 337: 464

[121] Sneddon AA, Wu HC, Farquharson A, Grant I, Arthur JR, Rotondo D,

Choe SN, Wahle KW (2003) Atherosclerosis 171: 57

[122] Nam SY, Baek IJ, Lee BJ, In CH, Jung EY, Yon JM, Ahn B, Kang JK, Yu

WJ, Yun YW (2003) J Reprod Dev 49: 389

[123] Lapointe J, Kimmins S, Maclaren LA, Bilodeau JF (2005) Endocrinology

146: 2583

[124] Roveri A, Casasco A, Maiorino M, Dalan P, Calligaro A, Ursini F (1992) J

Biol Chem 267: 6142

[125] Maiorino M, Wissing JB, Brigelius-Flohé R, Calabrese F, Roveri A,

Steinert P, Ursini F, Flohé L (1998) Faseb J 12: 1359

[126] Imai H, Hirao F, Sakamoto T, Sekine K, Mizukura Y, Saito M, Kitamoto T,

Hayasaka M, Hanaoka K, Nakagawa Y (2003) Biochem Biophys Res


24

Commun 305: 278

[127] Conrad M, Moreno SG, Sinowatz F, Ursini F, Kolle S, Roveri A,

Brielmeier M, Wurst W, Maiorino M, Bornkamm GW (2005) Mol Cell Biol

25: 7637

[128] Schneider M, Forster H, Boersma A, Seiler A, Wehnes H, Sinowatz F,

Neumuller C, Deutsch MJ, Walch A, Hrabe de Angelis M, Wurst W, Ursini

F, Roveri A, Maleszewski M, Maiorino M, Conrad M (2009) Faseb J 23:

3233

[129] Seiler A, Schneider M, Forster H, Roth S, Wirth EK, Culmsee C, Plesnila N,

Kremmer E, Radmark O, Wurst W, Bornkamm GW, Schweizer U, Conrad

M (2008) Cell Metab 8: 237

[130] Rapoport SM, Schewe T, Wiesner R, Halangk W, Ludwig P, Janicke-

Hohne M, Tannert C, Hiebsch C, Klatt D (1979) Eur J Biochem 96: 545

[131] Foresta C, Flohé L, Garolla A, Roveri A, Ursini F, Maiorino M (2002) Biol,

Reprod 67: 967

[132] Burk RF, Olson GE, Hill KE (2007) In: Hatfield DL, Berry MJ, Gladyshev

VN (eds) Selenium, Its Molecular Biology and Role in Human Health. 2nd

Ed., Springer, New York: 111

[133] Imai H, Koumura T, Nakajima R, Nomura K, Nakagawa Y (2003) Biochem

J 371: 799

[134] Nomura K, Imai H, Koumura T, Kobayashi T, Nakagawa Y (2000)

Biochem J 351: 183

[135] Brielmeier M, Bechet JM, Suppmann S, Conrad M, Laux G, Bornkamm

GW (2001) Biofactors 14: 179

[136] Brandes N, Schmitt S, Jakob U (2008) Antioxid Redox Signal 11: 997

[137] Cook HW, Lands WE (1976) Nature 260: 630

[138] Weitzel F, Wendel A (1993) J Biol Chem 268: 6288

[139] Imai H, Nakagawa Y (2003) Free Radic Biol Med 34: 145

[140] Heirman I, Ginneberge D, Brigelius-Flohé R, Hendrickx N, Agostinis P,

Brouckaert P, Rottiers P, Grooten J (2006) Free Radic Biol Med 40: 285

[141] Banning A, Brigelius-Flohé R (2005) Antioxid Redox Signal 7: 889

[142] Wenk J, Schuller J, Hinrichs C, Syrovets T, Azoitei N, Podda M, Wlaschek

M, Brenneisen P, Schneider LA, Sabiwalsky A, Peters T, Sulyok S,

Dissemond J, Schauen M, Krieg T, Wirth T, Simmet T, Scharffetter-

Kochanek K (2004) J Biol Chem 279: 45634

[143] Brigelius-Flohé R, Maurer S, Lotzer K, Böl G, Kallionpaa H, Lehtolainen P,

Viita H, Yla-Herttuala S (2000) Atherosclerosis 152: 307

[144] Vernet P, Rigaudiere N, Ghyselinck N, Dufaure JP, Drevet JR (1996)

Biochem Cell Biol 74: 125

[145] Drevet JR, Lareyre JJ, Schwaab V, Vernet P, Dufaure JP (1998) Mol

Reprod Dev 49: 131

[146] Chabory E, Damon C, Lenoir A, Kauselmann G, Kern H, Zevnik B, Garrel

C, Saez F, Cadet R, Henry-Berger J, Schoor M, Gottwald U, Habenicht U,

Drevet JR, Vernet P (2009) J Clin Invest 119: 2074

[147] Dear TN, Campbell K, Rabbitts TH (1991) Biochemistry 30: 10376

References

25

[148] Kryukov GV, Castellano S, Novoselov SV, Lobanov AV, Zehtab O, Guigo

R, Gladyshev VN (2003) Science 300: 1439

[149] Utomo A, Jiang X, Furuta S, Yun J, Levin DS, Wang YC, Desai KV, Green

JE, Chen PL, Lee WH (2004) J Biol Chem 279: 43522

[150] Rhee SG, Kang SW, Jeong W, Chang TS, Yang KS, Woo HA (2005) Curr

Opin Cell Biol 17: 183

Documents

[Advanced Topics in Science and Technology in China] Selenoproteins and Mimics Volume 703 || Glutathione Peroxidases