| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departamento de Estructura y Función de Proteínas, Instituto Reina Sofía de Investigaciones Nefrológicas, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas, 28006 Madrid, Spain; and
* Departamento de Bioquímica y Biología Molecular, Facultad de Veterinaria, Universidad de Córdoba, 14071 Córdoba, Spain
1Correspondence: Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas, Velázquez 144, 28006, Madrid, Spain. E-mail: pklatt@cib.csic.es (P.K.) or slamas{at}cib.csic.es (S.L.).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Key Words: glutathione • redox regulation • S • thiolation • transcription factor
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
, which represents the major low molecular weight
antioxidant in mammalian cells, plays a central role in the cellular
defense against oxidative damage (1)
. Apart from providing
the cell with a reducing environment and maintaining proteins in a
reduced state, accumulating evidence suggests that the glutathione
redox couple provides a means of dynamically regulating protein
function by reversible disulfide bond formation (2
3
4)
.
The formation of inter- and intramolecular disulfides as well as mixed
disulfides between protein cysteines and GSH, i.e., S-glutathiolation,
has been implicated in the stabilization of extracellular proteins,
protection of proteins against irreversible oxidation of critical
cysteine residues, and regulation of enzyme activity and transcription.
Redox regulation of transcription factors whose DNA binding activity is
reliant on the redox status of oxidant-sensitive cysteines in their
structures, such as nuclear factor-
B (NF-B), p53, SP-1, and
activator protein-1 (AP-1), was suggested as one mechanism by which
cells may transduce oxidative stress into the inducible expression of a
wide variety of genes implicated in cellular changes such as
proliferation, differentiation, and apoptosis (5
6
7)
.
However, the molecular mechanisms linking the cellular redox state to a
reversible oxidation of these DNA binding proteins are still poorly
understood.
In the present study, we have used the recombinant DNA binding domain
of the transcriptional activator c-Jun as a model protein to
investigate potential molecular mechanisms underlying the
posttranslational modification of cysteines in basic leucine zipper
(bZip) DNA binding proteins in response to changes in the ratio of
reduced (GSH) to oxidized (GSSG) glutathione. c-Jun is capable of
forming homodimers as well as heterodimers with proteins of the Fos
family. The dimeric Jun-Jun and Jun-Fos proteins compose the
sequence-specific transcriptional activator AP-1. Dimerization of these
proteins occurs via leucine zipper domains and serves to orientate the
monomers in a way such that a highly conserved zone rich in basic amino
acids forms a bipartite DNA binding site (8)
. The fact
that a conserved cysteine is mutated to serine in the oncogenic c-Jun
homologue v-Jun and that this point mutation is associated with loss of
redox control put forward the hypothesis that sulfhydryl oxidation may
be involved in the redox regulation of AP-1 (9
, 10)
.
Studies with truncated Fos and Jun constructs confirmed that redox
regulation of AP-1 in fact maps to a single conserved cysteine residue
located in the basic DNA binding site of AP-1 proteins (9
, 11)
. Although biochemical analysis indicates that the reduced
form of this cysteine is a prerequisite for DNA binding, the molecular
entity of the inactive and presumably oxidized cysteine residue is a
matter of ongoing debate and remains to be elucidated. Formation of an
intermolecular disulfide bond (11)
or the oxidation of the
cysteine to a sulfenate or sulfinate have been proposed as possible
mechanisms (9)
. However, a major drawback of the few
studies performed so far with purified AP-1 proteins, investigating the
redox regulation of DNA binding in in vitro models, is that
they were performed in the absence of GSH, neglecting the presence of
millimolar concentrations of the glutathione redox couple in
vivo. To address this issue, we analyzed purified recombinant
c-Jun DNA binding domains for redox-dependent modifications of
conserved cysteine residues and concomitant changes in DNA binding. In
our experimental system, GSH/GSSG mixtures were used to provide the
potential to oxidize the sulfhydryls contained in the DNA binding
module of c-Jun. To assign the observed effects to one of the two
cysteines located in the DNA binding site (Cys 269) and close to the
leucine zipper of the transcription factor (Cys 320), data obtained
with the truncated wild-type construct were compared with those from
the respective cysteine to serine mutants (see Scheme 8).We provide evidence that a decrease in the ratio of reduced to oxidized
glutathione induces S-glutathiolation of the cysteine located in the
DNA binding site of c-Jun as well as formation of an intermolecular
disulfide bridge between cysteines close to the leucine zipper of the
protein. We show that S-glutathiolation may reversibly redox-regulate
c-Jun DNA binding in our system; based on a molecular model of the
S-glutathiolated c-Jun homodimer, we discuss the structural elements
facilitating S-glutathiolation of this transcription factor.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
, 13)
.
Tritium-labeled glutathione ([3H]GSH, 45–50
Ci/mmol, ~0.02 mM) was from Dupont NEN (Frankfurt/Main, Germany) and
adjusted to a final concentration of 10 mM by the addition of 10
volumes of an 11 mM solution of unlabeled GSH (Sigma, St. Louis, Mo.).
[3H]GSSG was prepared by oxidation of
[3H]GSH as described (14)
. GSH
(free acid, SigmaUltra), GSSG (free acid, SigmaUltra), and
5,5'-thiobis(2-nitrobenzoate) were purchased from Sigma-Aldrich
(Milwaukee, Wis.). Boehringer Mannheim (Mannheim, Germany) provided
yeast glutathione reductase (120 U/mg).
Determination of GSH and GSSG concentrations
GSH concentrations were determined photometrically by Ellman's
assay (15)
, which was modified as described recently
(16)
, measuring the GSH-dependent formation of
5-thio-2-nitrobenzoate (
412 nm = 13.6
mM-1 cm-1) from
5,5'-thiobis(2-nitrobenzoate). GSSG concentrations were determined by a
coupled assay (17)
as glutathione reductase-mediated
reduction of GSSG at the expense of NADPH (340
nm = 6.34 mM-1
cm-1).
Preparation of recombinant c-Jun DNA binding domains
Expression and purification of recombinant c-Jun fragments were
performed similar to a published procedure (9
, 18)
.
Because of numerous modifications, the method is described briefly
below.
The insert coding for the DNA binding domain of human c-Jun, corresponding to amino acids 223–327 of the translated sequence (GenBank accession number J04111), was amplified by polymerase chain reaction (PCR) from a clone containing the entire coding sequence of human c-Jun, using the forward and reverse PCR primers 5'-CGC GGA TCC CAG GCC CTG AAG GAG GAG-3' and 5'-GAG GGA AGC TTA CTG CTG CGT TAG CAT GAG TT-3', respectively. The obtained PCR fragment was digested with BamHI/HindIII and ligated into the BamHI-HindIII site of the hexahistidine fusion protein expression vector pQE-30 (Qiagen, Chatsworth, Calif.). The protein derived from the resulting insert, which contains one cysteine in the basic DNA binding site (Cys 269) and a second cysteine close to the leucine zipper (Cys 320) was designated as CC-Jun (wild-type human c-Jun DNA binding domain).
A mutant c-Jun DNA binding domain in which cysteine 269 was replaced by a serine (the corresponding protein was designated as SC-Jun) was constructed by PCR using the BamHI-HindIII CC-Jun insert described above as template. A 5'-fragment of SC-Jun was amplified, using the forward primer described for the wild-type construct and the mutant reverse primer 5'-CTT CCT TTT TCG CGA CTT GGA GGC AGC-3'. The 3' fragment coding for SC-jun was obtained in an analogous way, using the complementary mutant forward primer 5'-GCT GCC TCC AAG TCG CGA AAA AGG AAG-3' and the wild-type reverse primer described above. The mutant primers introduced a unique NruI site into the SC-Jun constructs due to the cysteine to serine mutation. The obtained 5'and 3' PCR fragments were digested with BamHI/NruI and NruI/HindIII, respectively, and cloned into the BamHI-HindIII site of the expression vector pQE-30.
A mutant c-Jun DNA binding domain, in which cysteine 320 was replaced by a serine (the corresponding protein was designated as CS-Jun), was constructed by PCR using the CC-Jun BamHI-HindIII insert described above as template. The forward primer was the same as described above for CC-Jun; the mutant reverse primer, which introduced a unique AvaII site into CS-Jun due to the cysteine to serine mutation, was 5'-GAG GGA AGC TTA CTG CTG CGT TAG CAT GAG TTG GGA-3'. The PCR product was digested with BamHI/HindIII and ligated into the BamHI-HindIII site of the expression vector pQE-30.
The resulting plasmids were transformed into competent
Escherichia coli (M15[pRep4], Qiagen) according to the
instructions of the manufacturer. Recombinant clones coding for the DNA
binding domains of c-Jun were verified by restriction analysis,
dideoxynucleotide sequencing, and their ability to direct expression of
the respective hexahistidine fusion protein. Cysteine to serine
mutations were further confirmed by determinations of the cysteine
content of the purified proteins using the modified Ellman's assay
(16)
. For the expression of c-Jun proteins, 1 l of modified LB medium, containing 25 g bacto-tryptone, 15 g
bacto-yeast extract, 5 g NaCl, 100 mg ampicillin, and 25 mg
kanamycin were inoculated with 15 ml of an overnight starter culture of
the recombinant clones. Bacteria were grown at 37°C to an optical
density of 0.6–0.8 (600 nm) and induced with 1 mM
isopropyl-ß,D-thiogalactopyranoside. After incubating the culture for
an additional 4 h, bacteria were harvested by centrifugation at
4°C for 30 min at 5000 x g. The obtained cell
pellet was resuspended in 50 ml of a 25 mM phosphate buffer (pH 8.0),
containing 6 M guanidine hydrochloride, 0.1% (v/v) 2-mercaptoethanol
(buffer A), and frozen at -80°C. The frozen cell suspension was
thawed rapidly at 37°C and stirred for 1 h at room temperature.
After centrifugation of the cell lysate at 20,000 x
g for 30 min, 3–4 ml (bed volume) of nickel-chelate
resin (Qiagen) equilibrated with buffer A was added to the obtained
supernatant and the suspension was stirred for another 2 h at room
temperature. The mixture was poured into a chromatography column (inner
diameter 1.5 cm) and washed with 15 bed volumes of buffer A, followed
by 10 bed volumes of buffer A adjusted to pH 6.5 with HCl and 0.3 bed
volumes of buffer A adjusted to pH 4.5 with HCl. Finally, the protein
was eluted with 5 bed volumes of buffer A adjusted to pH 4.5. The
eluate was dialyzed for 4 h at 4°C against 1 l of a 25 mM
phosphate buffer (pH 7.4) containing 1 mM EDTA, 5% (v/v) glycerol,
0.1% (v/v) 2-mercaptoethanol, 0.01% (v/v) Nonidet P-40 (buffer B),
and 2.5 M guanidine hydrochloride. To reduce the concentration of
guanidine hydrochloride, the dialysis buffer was diluted stepwise 1:2
(twice) and 1:5 (once) with fresh buffer B and dialyzed for another
4–12 h after each dilution step. Finally, dialysis was continued for
24 h against four changes of 1 l of buffer B. The total time
of dialysis was at least 48 h. The dialyzed protein was
concentrated to a final volume of 6–8 ml with Vivapore 20
concentrators (Vivascience, Chartres, France). The purity of
the obtained protein preparations was estimated >95% as judged by
Coomassie-stained sodium dodecyl sulfate gels. Concentrations of the
purified proteins were determined by amino acid analysis and were in
the range of 0.4–0.5 mM.
Analysis of c-Jun DNA binding by EMSA
c-Jun DNA binding domains were preincubated at a final
concentration of 10 µM in a 20 mM Tris/HCl buffer (pH 7.5) containing
50 mM NaCl, 5 mM MgCl2, 1 mM EDTA, 5% (v/v)
glycerol, and 0.01% (v/v) Nonidet P-40 (buffer C) in the presence of
GSH/GSSG, as described in the respective figure legends. To determine
DNA binding activity, 2 µl aliquots of the preincubation mixture were
diluted into 16 µl of the same redox buffer, i.e., maintaining the
GSH/GSSG ratio of the preincubation, which also contained 0.2 mg/ml
bovine serum albumin to stabilize the diluted Jun protein and 0.1 mg/ml
poly(dI-dC) to block nonspecific DNA binding. Finally, 2 µl of the
32P-radiolabeled double-stranded AP-1
oligonucleotide 5'-GGG CTT GAT GAG TCA GCC GGA CC-3' was added. The
samples were incubated for an additional 30 min at room temperature to
permit DNA binding prior to electrophoresis at 200 V on
pre-electrophoresed 6% nondenaturing polyacrylamide gels with 22 mM
Tris borate/0.5 mM EDTA as running buffer. Gels were dried, visualized
by autoradiography, and analyzed by densitometry. To study the
reversibility of c-Jun inactivation by S-glutathiolation, c-Jun DNA
binding domains (10 µM) were S-thiolated by preincubation for 30 min
at 37°C in the presence of 1 mM GSSG as described above.
Subsequently, 2 µl of the preincubation mixture was diluted in the
absence and presence of 10 mM GSH into 20 µl of buffer C containing
0.2 mg/ml bovine serum, 0.1 mg/ml poly(dI-dC), and the
32P-labeled oligonucleotide. Samples were
incubated for 60 min at 37°C, cooled to room temperature, and
subjected to electrophoresis on nondenaturing gels as described above.
Analysis of disulfide bond formation between c-Jun subunits
c-Jun DNA binding domains were incubated at a final
concentration of 10 µM in buffer C under the conditions indicated in
the text. Reactions were stopped by the addition of iodoacetamide at a
final concentration of 50 mM and incubation for 30 min at room
temperature. Where indicated, proteins were incubated for an additional
30 min in the presence of 10 mM DTT at room temperature prior to the
addition of iodoacetamide. Samples (4 µg c-Jun) were subjected to
nonreducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) on discontinuous 1 mm slab gels (7 x 8 cm), which
contained acrylamide and bisacrylamide at final concentrations of 16
and 0.1% (w/v), respectively. Gels were stained for protein with
Coomassie blue R-250, dried, and analyzed by densitometry.
Analysis of c-Jun glutathiolation
c-Jun DNA binding domains (10 µM) were incubated in a final
volume of 0.1 ml for the indicated times at 37°C in buffer C in the
presence of 3H-labeled GSH and GSSG at the
concentrations given in the text. To isolate the glutathiolated protein
by trichloroacetic acid precipitation according to a published method
(14)
, reactions were stopped by the addition of 0.9 ml
ice-cold trichloroacetic acid and incubation on ice for 30 min. Samples
were centrifuged for 10 min at 20,000 g and the supernatant was
discarded. Subsequent to washing the precipitated protein three times
with 0.9 ml ice-cold trichloroacetic acid, the protein pellet was
dissolved by treatment with 0.1 ml of 0.5 N NaOH for 20 min at 70°C
and assayed for incorporation of [3H]GSH by
liquid scintillation counting. Results were corrected for protein
recovery, which under these conditions was 68 ± 7%
(n=6), and for blank values, which were determined as
non-DTT-releasable radiolabel by treating the radiolabeled proteins for
60 min at 37°C with 10 mM DTT prior to trichloroacetic acid
precipitation. To study the reversibility of Jun S-glutathiolation,
c-Jun DNA binding domains (10 µM) were S-thiolated by preincubation
for 30 min at 37°C in the presence of 1 mM
[3H]GSSG as described above. Subsequently, 0.1
ml of the preincubation mixture, containing the thiolated protein, was
diluted into 1 ml of buffer C in the absence and presence of 10 mM GSH.
In some experiments, 1.4 µM purified human glutaredoxin or 2 µM of
purified human thioredoxin plus an excess of partially purified rat
liver thioredoxin reductase and 0.1 mM NADPH were also present. After
incubation for 1 h at 37°C, samples were precipitated by the
subsequent addition of 0.1 ml of 30 mg/ml bovine serum albumin and 0.12
ml 100% (w/v) trichloroacetic acid. The amount of incorporated
[3H]GSH was determined as described above and
corrected for blank values determined in the presence of 10 mM DTT.
Enzymatic activities of glutaredoxin and thioredoxin were confirmed
under the experimental conditions used in this study with hydroxyethyl
disulfide (19)
and insulin disulfide (20)
as
model substrates, respectively. Under assay conditions, the specific
activities of glutaredoxin and thioredoxin were 30 and 24 U/mg,
respectively.
Molecular modeling
3-Dimensional models of S-glutathiolated c-Jun homodimers and
c-Jun/cFos heterodimers were built from the previously published 3.0
Å-resolution X-ray crystal structure of the bZip region of c-Jun/cFos
(8)
and the nuclear magnetic resonance (NMR) solution
structure of the leucine zipper domain of the c-Jun homodimer
(21)
. Their cartesian coordinates were from the Brookhaven
Protein Data Bank (PDB codes 1FOS and 1JUN). For the simulation of
S-glutathiolation, GSH molecules were randomly inserted into the
protein via disulfide bridges to target cysteines; the overall protein
structure was subjected to energy minimization until convergence, using
a combination of steepest descents and conjugate gradients algorithms.
The energy calculations were carried out under the AMBER force field
(22)
. Computations were performed on a Power Challenge
R10000 by using the BIOSYM software package, release 95.0 (Molecular
Simulations, Inc., San Diego, CA).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
. In the
resting cell, the ratio of GSH to GSSG exceeds 100, whereas in various
models of oxidative stress this ratio was reported to decrease to
values between 10 and 1 (2)
. To find out whether DNA
binding activity of c-Jun is modulated by changes in the GSH/GSSG
ratio, we pre-equilibrated purified c-Jun DNA binding modules in
GSH/GSSG redox buffers before determining its DNA binding activity by
electrophoretic mobility shift assays (EMSA). GSH/GSSG ratios were
expressed as the molar ratios of glutathione equivalents, i.e., the
proportion of GSH to 2 GSSG, and refer to ratios adjusted at the
beginning of the experiment. During the incubation period of 1 h,
GSH suffered some autoxidation, which resulted in a shift of the
initial GSH/2 GSSG ratios of 100, 10, 1, and 0.1 to 61 ± 7,
8.5 ± 0.5, 1.0 ± 0.1, and 0.08 ± 0.01
(n=3, t = 1 h), respectively. Figure 1
A shows that a decrease of the ratio of reduced to oxidized
glutathione induced a pronounced inhibition of c-Jun DNA binding
activity. Half-maximal (51 ± 9%), 83 ± 10%, and 98 ± 2% inhibitions were observed at GSH/2 GSSG ratios of 10, 1, and
0.1, respectively (mean ± SE,
n=4). The observed inactivation was entirely reversed by the
dithiol DTT, confirming the involvement of a reversible protein
sulfhydryl modification in the GSH/GSSG-mediated redox control of c-Jun
DNA binding.
|
The CC-Jun homodimer contains two pairs of cysteines in its DNA binding
domain (see Scheme 8
), one located in the basic region that makes
direct contact with DNA (Cys 269) and one upstream of the adjacent
leucine zipper (Cys 320). A comparison of cysteine to serine mutants
with the wild-type (CC) construct shows that the mutant lacking the
cysteine in its DNA binding site (SC) escaped glutathione-mediated
redox control, whereas the mutant lacking the cysteine in its leucine
zipper (CS) remained susceptible to oxidative inactivation (Fig. 1B
). Under reducing conditions, DNA binding activities of
the wild-type and mutant c-Jun proteins were virtually identical (not
shown). To find out whether this targeted GSH/GSSG-dependent
down-regulation of DNA binding activity is a reversible process, we
inactivated CC-Jun by preincubation with 1 mM GSSG prior to diluting
the protein 1:10 in the absence and presence of 10 mM GSH and
determining its DNA binding activity (Fig. 1C
). In the
presence of GSH, DNA binding activity of CC-Jun was restored to 
70% as compared with controls, which had been maintained under
reducing conditions throughout the total incubation period. These data,
therefore, indicate that the glutathione redox couple may regulate DNA
binding of c-Jun in a reversible manner by modifying a single conserved
cysteine residue in the DNA binding site of the transcription factor.
Redox-dependent formation of intermolecular disulfide bridges
between c-Jun subunits
A shift of the redox potential to oxidative conditions may
induce the formation of intermolecular disulfide bonds between vicinal
cysteine residues of protein subunits. Formation of disulfide bonds
between the monomers of Jun and Fos proteins via oxidation of a
cysteine in their basic DNA binding sites has been proposed by one
group to account for redox control of AP-1 DNA binding
(11)
, whereas others provided experimental evidence
against this mechanism (9)
. To address this issue in our
redox model, we pre-equilibrated CC-Jun in GSH/GSSG buffers before
resolving monomers from covalently linked dimers by nonreducing
SDS-PAGE (Fig. 2
A). Under reducing conditions, i.e., at a GSH/2 GSSG ratio of
100, more than 90% of CC-Jun migrated as monomers, whereas a decrease
of the GSH/GSSG ratio induced the formation of SDS-resistant
homodimers. Half-maximal dimerization was observed at GSH/2 GSSG ratios
between 1 and 0.1, and ratios of 0.001 (not shown) were required to
convert >90% of CC-Jun into covalently linked dimers. The
reversibility of this effect by DTT suggests the formation of an
intermolecular disulfide bridge between one or both of the two
cysteines contained in the CC-Jun monomer. A comparison of cysteine to
serine mutants (Fig. 2B
) shows that SC-Jun but not CS-Jun
was converted into disulfide-linked dimers. Even in the virtual absence
of GSH, GSSG (3 mM) did not induce any detectable dimerization of
CS-Jun (not shown), demonstrating that redox-regulated disulfide
formation is specifically targeted to the cysteine located in the
leucine zipper of CC-Jun. As evident from functional assays (see Fig. 1B
), however, this cysteine residue is apparently not
involved in the GSH/GSSG dependence of CC-Jun DNA binding, thus
excluding disulfide formation as a mechanism for the redox control of
c-Jun DNA binding activity in our system.
|
Glutathiolation of a conserved cysteine residue in the DNA binding
site of c-Jun
Pioneering work by Thomas and co-workers (3)
,
providing evidence that some cellular proteins form mixed disulfides
with glutathione in response to oxidative stress, has introduced the
concept of regulation of protein function by S-glutathiolation into the
field of redox biochemistry. Studies with purified proteins have
established various enzymes, including glucose-6-phosphate
dehydrogenase, phosphofructokinase, aldose reductase, fatty acid
synthase, carbonic anhydrase, hydroxymethylglutaryl-CoA reductase,
glutathione transferase, and HIV-1 protease, as potential targets for
redox-dependent S-glutathiolation (reviewed in refs 3
, 4
, 24
, 25
). Recently, ubiquitin-conjugating enzymes (26
, 27)
have been added to the list of proteins regulated by
S-thiolation. The observation that oxidation-inactivated nuclear factor
1 can be reactivated by the thioltransferase glutaredoxin
(28)
further provides a first indication that the
mechanism of S-glutathiolation might also be a valid model for the
redox regulation of transcription.
To investigate whether S-glutathiolation may account for the redox
control of c-Jun DNA binding, we equilibrated purified c-Jun DNA
binding modules in redox buffers, containing tritium-labeled GSH and
GSSG, and subsequently isolated the
S-[3H]glutathiolated protein by trichloroacetic
acid precipitation. At a [3H]GSH/2
[3H]GSSG ratio of 1, i.e., under conditions
that almost completely abolished DNA binding of CC-Jun (compare Fig. 1A
), [3H]GSH was incorporated into
the transcription factor in a time-dependent manner with an apparent
half-time of ~2 min and the stoichiometric incorporation of ~1 mol
of GSH per mole of protein monomer at t 30 min (Fig. 3
A, filled symbols). Incorporation of the radiolabel was
almost entirely reversed by DTT (release of >85% of
[3H]GSH incorporated at a GSH/2 GSSG ratio of
0.1; not shown), clearly indicating that the incorporation of
glutathione occurred via the formation of a disulfide bond between the
protein and the tripeptide. Control incubations in the virtual absence
of [3H]GSSG, i.e., in the presence of 3 mM
[3H]GSH (Fig. 3A
, open symbols),
revealed a time-independent and negligible incorporation of less than
0.01 mol of the radiolabel per mole protein, indicating that GSH
autoxidation was not involved in the time-dependent formation of a
mixed disulfide. As shown in Fig. 3B
, mutation of the
cysteine residue in the DNA binding site of c-Jun to serine (SC-Jun)
largely abolished glutathiolation (0.06 ± 0.01 mol
[3H]GSH per mole protein), whereas a cysteine
to serine mutation in the leucine zipper module of the protein (CS-Jun)
virtually did not affect the degree of glutathiolation (1.10 ±
0.04 mol of [3H]GSH per mole protein) as
compared with CC-Jun (1.09 ± 0.06 mol of
[3H]GSH per mole protein). These data
demonstrate that the formation of a mixed disulfide with glutathione is
targeted to the same cysteine residue, which apparently accounts for
the redox sensitivity of c-Jun DNA binding (see Fig. 1B
).
|
Reversible regulation of c-Jun S-glutathiolation by the glutathione
redox couple
Figure 3C
shows the dependency of CC-Jun S
glutathiolation on the ratio of reduced to oxidized glutathione. At a
[3H]GSH/2 [3H]GSSG
ratio of 100, we did not detect any significant levels of
S-[3H]glutathiolation, whereas at ratios of 10
and 1 the incorporation of [3H]GSH increased to
0.43 ± 0.02 and 1.09 ± 0.06 mol per mole protein monomer,
respectively. The incorporation of
[3H]glutathione into CC-Jun was only slightly
increased by further lowering the [3H]GSH/2
[3H]GSSG ratio by one order of magnitude to 0.1
(1.19 ± 0.04 mol per mole protein monomer), suggesting a maximal
incorporation of 1 mol glutathione per mole of CC-Jun monomer. From
these data and dose response experiments, varying GSH/GSSG ratios and
total glutathione concentrations (not shown), the ratio of GSH to 2
GSSG giving half-maximal S-glutathiolation
(Kmix), was estimated to be 13, which corresponds
to a calculated (2)
standard redox potential of
approximately -0.27 V.
To find out whether the observed S-glutathiolation of c-Jun can be reversed by restoring reducing conditions, we S-glutathiolated CC-Jun by preincubation with 1 mM [3H]GSSG prior to inducing dethiolation by a 1:10 dilution of the protein in the absence and presence of 10 mM GSH by raising the GSH/2 GSSG ratio to 50. At the beginning of the dethiolation reaction (t = 0), 0.93 ± 0.13 mol (mean ± SE, n=4) per mole CC-Jun monomer were found to be S-glutathiolated. After incubation at 37°C for 60 min in the absence of GSH, no significant loss of the radiolabel was observed, whereas in the presence of the reduced sulfhydryl c-Jun was dethiolated to ~50% (0.45 ± 0.05 mol of [3H]GSH per mole protein monomer; mean ± SE, n=4), confirming redox reversibility of c-Jun glutathiolation.
Given the low velocity of nonenzymatic glutathione removal, it is
attractive to speculate that enzymatic mechanisms may be implicated in
a rapid dethiolation of the transcription factor. While glutaredoxin
appears to display some specificity for mixed disulfides, thioredoxin
has been proposed to preferentially catalyze the cleavage of
intramolecular protein disulfides (29)
. On the other hand,
the thioredoxin/thioredoxin reductase system was shown to reactivate
oxidatively inactivated Jun (30)
, and thus may be a
possible candidate to mediate c-Jun dethiolation. Under our
experimental conditions, however, neither glutaredoxin nor the
thioredoxin/thioredoxin reductase system catalyzed any significant
dethiolation of CC-Jun (data not shown). This may be explained by lack
of accessibility of the modified cysteines (which are located at the
inner side of the bipartite, positively charged DNA binding site) to
these >10 kDa proteins due to steric or electrostatic constraints. It
remains to be investigated whether other as yet unidentified
disulfide-reducing or -transferring enzymes are involved in the
dethiolation of c-Jun. One candidate may be the DNA repair protein
redox factor 1, which stimulates the transcriptional activity of AP-1,
presumably by a redox mechanism involving tightly associated
thioredoxin (31
, 32)
.
Specific c-Jun-glutathione interactions
There is neither conclusive evidence for enzymatic catalysis nor a
generalized nonenzymatic mechanism of protein S-glutathiolation. It has
been proposed that mixed disulfide formation occurs in response to
changes in the GSH/GSSG ratio through a thiol/disulfide exchange
mechanism (2)
. This mechanism may be relevant in
situations of oxidative stress that are associated with a pronounced
change in the ratio of reduced to oxidized glutathione. However,
although specific glutathione–protein interactions have been suggested
to shift the protein-mixed disulfide equilibrium toward the thiolated
species, the equilibrium constants (Kmix) found
for most proteins are usually ~1 (2)
. From the results
obtained in this study, we estimate the Kmix for
c-Jun to be 13, indicating that the transcription factor might in fact
be S-glutathiolated in response to changes in the GSH/GSSG ratios as
they have been described in models of oxidative stress (i.e., 100 > GSH/2 GSSG > 1). An alternative mechanism for redox-dependent
protein S-thiolation is based on the intriguing observation that mixed
disulfide formation may occur in intact cells in response to oxidative
stress without any detectable changes in the GSH/GSSG ratio (33
, 34)
. This phenomenon has been rationalized by a direct oxidation
of protein sulfhydryls that precedes the covalent addition of reduced
glutathione to the protein. As discussed in depth in a recent review on
this topic (3)
, this mechanism implicates that
S-thiolation sites bind reduced glutathione and thereby facilitate the
formation of a mixed disulfide with GSH after an initial oxidative
activation event that generates a reactive protein sulfhydryl
intermediate such as a thiol radical or a sulfenic acid. Of note, both
models for mixed disulfide formation postulate specific
protein–glutathione interactions as one of the key factors determining
the susceptibility of a protein to redox-dependent S-glutathiolation.
To shed some light on the structural elements that render the DNA
binding domain of c-Jun susceptible to S-glutathiolation, we
constructed a structural model of the S-glutathiolated c-Jun homodimer
based on previously published X-ray and NMR data of the unmodified
protein (8
, 21)
. In our energy-minimized model of the
S-glutathiolated c-Jun DNA binding site (Fig. 4
, one GSH molecule was bound in an energetically favorable extended
conformation to each subunit of the transcription factor homodimer via
disulfide bridges to cysteine residues 269. Incorporation of the
tripeptides in such an energetically stable, crossed orientation
neither caused any significant conformational changes in the
tridimensional arable part of the DNA contact interface, and thus are
expected to seriously interfere with DNA binding. This conclusion would
be in accordance with our observation that S-glutathiolation is
associated with a loss of c-Jun DNA binding activity.
|
As discussed, the exceptionally high susceptibility of c-Jun to mixed
disulfide formation may be explained by specific protein–GSH
interactions. Our model suggests that the negatively charged GSH
tripeptide may in fact be bound and stabilized by electrostatic
interactions with the positively charged DNA binding site of the
transcription factor homodimer (Fig. 4B
). The two negatively
charged oxygen atoms of the terminal 
-carboxylate groups of the
-glutamyl and glycyl moieties of the GSH molecule, which are bound
to cysteine 269 of c-Jun (right chain), are stabilized by salt bridges
to the positively charged ammonium groups of arginine 270 (2.4 Å) and
lysine 273 (2.8 Å) of the same c-Jun subunit, respectively. Similarly,
the second GSH molecule bound to cysteine 269 of the second c-Jun
subunit (left chain) was stabilized by a salt bridge between the
-glutamyl carboxylate group of GSH and arginine 270 (1.8 Å) of the
same subunit, whereas the second salt bridge, fixing the glycyl
carboxylate moiety of GSH to the protein, was established to arginine
276 (1.8 Å) of the neighboring c-Jun polypeptide. We propose that the
c-Jun structure may facilitate the formation of a mixed disulfide with
GSH by binding and stabilizing the tripeptide in an extended
conformation, fixing the negatively charged terminal carboxylate
moieties of the tripeptide via salt bridges to positively charged
basic amino acid residues located in the DNA binding site of both
polypeptide chains of the transcription factor homodimer. In a
molecular model of the S-glutathiolated cFos/c-Jun heterodimer (not
shown), a virtually identical pattern of salt bridges, involving
equivalent arginine and lysine residues in cFos, was found to stabilize
GSH in the mixed disulfide.
It has been suggested that proteins, which specifically recruit GSH as
a redox agent, display a recognizable electrostatic pattern, fixing the
sulfhydryl to the protein in an extended conformation by direct salt
bridges or hydrogen bonds to the terminal -carboxylate groups of its
-glutamyl and glycyl moieties (3)
. These minimal
structural requirements for specific GSH binding have been derived from
crystal structure and NMR analysis of GSH binding proteins, including
the family of glutathione S-transferases (35)
,
glutaredoxin (36)
, and glutathione peroxidase
(37)
. Our molecular model of S-glutathiolated c-Jun
predicts that the DNA binding site of this transcription factor in fact
displays such a `GSH-footprint', and thus may provide an explanation
of why c-Jun represents such an exceptionally good target for
S-glutathiolation in vitro. Furthermore, it is tempting to
speculate that these modeled ionic interactions of GSH with neighboring
basic residues of the protein make enzymatic dethiolation of c-Jun by
thioredoxin and glutaredoxin unfavorable. Similarly, glutaredoxin was
found to be incapable of readily removing glutathione from cysteine 67
of glutathiolated HIV protease, which is also surrounded by basic amino
acids (38)
. Definite insights into the structure of the
mixed disulfide, which allow the classification of c-Jun as a GSH
binding protein, however, await X-ray crystallographic analysis of the
thiolated transcription factor.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Furthermore, with the DNA binding site of c-Jun, we have
identified an oxidant sensitive cysteine residue flanked by basic amino
acids as a novel target for S-glutathiolation. The structural model
presented in this study suggests that the structural motif, exemplified
by the DNA binding site of c-Jun, specifically recruits GSH as a redox
reagent by stabilizing the protein/GSH adduct via specific
intermolecular interactions. In terms of structural properties, this
motif is common to a number of cysteine-containing transcription
factors, including members of the Jun/Fos and ATF/CREB family as well
as c-Myb and NF-B. The unprecedented phenomenon of redox-triggered
S-thiolation of a transcription factor, therefore, may not only be
valid for c-Jun, but may represent a novel and general mechanism by
which transcription factors integrate sulfhydryl redox chemistry into a
functional response to oxidative stress.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Abbreviations: bZip, basic leucine zipper; GSH,
glutathione; GSSG, glutathione disulfide; DTT, dithiothreitol, AP-1,
activator protein-1; CC-Jun, wild-type human c-Jun DNA binding domain;
NMR, nuclear magnetic resonance; SC-Jun, human c-Jun DNA binding domain
with a cysteine 269 to serine mutation; CS-Jun, human c-Jun DNA binding
domain with a cysteine 320 to serine mutation; NF-B, nuclear
factor-B; PCR, polymerase chain reaction; SDS-PAGE, sodium dodecyl
sulfate-polyacrylamide gel electrophoresis; EMSA, electrophoretic
mobility shift assay.
Received for publication December 14, 1998. Revised for publication March 29, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Hochgrafe, J. Mostertz, D.-C. Pother, D. Becher, J. D. Helmann, and M. Hecker S-Cysteinylation Is a General Mechanism for Thiol Protection of Bacillus subtilis Proteins after Oxidative Stress J. Biol. Chem., September 7, 2007; 282(36): 25981 - 25985. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Koike, T. Hisada, M. Utsugi, T. Ishizuka, Y. Shimizu, A. Ono, Y. Murata, J. Hamuro, M. Mori, and K. Dobashi Glutathione Redox Regulates Airway Hyperresponsiveness and Airway Inflammation in Mice Am. J. Respir. Cell Mol. Biol., September 1, 2007; 37(3): 322 - 329. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Martinez-Ruiz and S. Lamas Signalling by NO-induced protein S-nitrosylation and S-glutathionylation: Convergences and divergences Cardiovasc Res, July 15, 2007; 75(2): 220 - 228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Qanungo, D. W. Starke, H. V. Pai, J. J. Mieyal, and A.-L. Nieminen Glutathione Supplementation Potentiates Hypoxic Apoptosis by S-Glutathionylation of p65-NF{kappa}B J. Biol. Chem., June 22, 2007; 282(25): 18427 - 18436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. F. Baker, T. Landowski, R. Dorr, W. R. Tate, J. M.C. Gard, B. E. Tavenner, T. Dragovich, A. Coon, and G. Powis The Antitumor Agent Imexon Activates Antioxidant Gene Expression: Evidence for an Oxidative Stress Response Clin. Cancer Res., June 1, 2007; 13(11): 3388 - 3394. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Berndt, C. H. Lillig, and A. Holmgren Thiol-based mechanisms of the thioredoxin and glutaredoxin systems: implications for diseases in the cardiovascular system Am J Physiol Heart Circ Physiol, March 1, 2007; 292(3): H1227 - H1236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Sykes, A. L. Mowbray, and H. Jo Reversible Glutathiolation of Caspase-3 by Glutaredoxin as a Novel Redox Signaling Mechanism in Tumor Necrosis Factor-{alpha}-Induced Cell Death Circ. Res., February 2, 2007; 100(2): 152 - 154. [Full Text] [PDF]
|
|
|
|
 |
|
 J. GAYARRE, M. ISABEL AVELLANO, F. J. SANCHEZ-GOMEZ, M. J. CARRASCO, F. J. CANADA, and D. PEREZ-SALA Modification of Proteins by Cyclopentenone Prostaglandins is Differentially Modulated by GSH in Vitro Ann. N.Y. Acad. Sci., January 1, 2007; 1096(1): 78 - 85. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. L. Reynaert, A. van der Vliet, A. S. Guala, T. McGovern, M. Hristova, C. Pantano, N. H. Heintz, J. Heim, Y.-S. Ho, D. E. Matthews, et al. Dynamic redox control of NF-{kappa}B through glutaredoxin-regulated S-glutathionylation of inhibitory {kappa}B kinase beta PNAS, August 29, 2006; 103(35): 13086 - 13091. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Rahman and I. M. Adcock Oxidative stress and redox regulation of lung inflammation in COPD. Eur. Respir. J., July 1, 2006; 28(1): 219 - 242. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
