© 1998 American Heart Association, Inc.
Original Contributions |
Inhibition of Tumor Necrosis Factor-
– and Interleukin-1–Induced Endothelial E-Selectin Expression by Thiol-Modifying Agents
From the German Institute of Human Nutrition Potsdam-Rehbrücke (B.F., R.B.-F.) and the Institute of Nutritional Science, University of Potsdam (C.M., R.B.-F.), Potsdam-Rehbrücke, Germany.
Correspondence to Dr Regina Brigelius-Flohé, German Institute of Human Nutrition Potsdam-Rehbrücke, Arthur-Scheunert-Allee 114-116, D-14558 Bergholz-Rehbrücke, Germany. E-mail flohe{at}www.dife.de
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—The expression of endothelial-leukocyte adhesion molecules has been postulated to be regulated by redox-sensitive events. Tumor necrosis factor-
(TNF-)– and interleukin-1 (IL-1)–induced E-selectin
expression was analyzed after pretreating human umbilical vein
endothelial cells with different thiol-modifying
agents, ie, diamide, phenylarsine oxide,
N-ethylmaleimide, and diethyl maleate. E-selectin
protein expression was quantified by indirect
immunofluorescence. All compounds suppressed the
cytokine-induced E-selectin expression in a
concentration-dependent manner, whereas the antioxidant
N-acetylcysteine showed no effect. The
inhibitory effect of diamide (100 µmol/L, 1 hour)
was reversible within 6 hours when the cells were allowed to recover
before application of cytokines. Reversibility was strongly
delayed when cells were deprived of glutathione by buthionine
sulfoximine pretreatment. Glutathione depletion alone did not influence
cytokine-induced E-selectin expression. Analysis of
cellular glutathione status showed a 3-fold increase in oxidized
glutathione after diamide treatment. Monochlorobimane labeling also
revealed a decrease in total cellular thiols. During recovery, the
glutathione status was restored within 1 hour, whereas total thiol
content and E-selectin expression needed at least 6 hours to return to
baseline. Complete inhibition of E-selectin expression by the vicinal
thiol blocker phenylarsine oxide (0.5 µmol/L) was reversed by
dithiols like dithiothreitol or dimercaptopropanol, but not by the
monothiol 2-mercaptoethanol. These data suggest that proteins with
essential thiols, most probably vicinal thiols. are involved in the
IL-1– and TNF-–mediated induction of E-selectin. These thiols must
be in the reduced state; oxidation or other modification thereof
attenuates or abolishes the cells' response to the cytokines.
Key Words: E-selectin • tumor necrosis factor- • interleukin-1 • diamide • phenylarsine oxide
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Endothelial E-selectin, first described as ELAM-1 (endothelial-leukocyte adhesion molecule-1),1 belongs to the selectin family of cellular adhesion molecules (CAMs) mediating the initial attachment of leukocytes to vascular endothelial cells in early inflammatory events that may ultimately lead to atherosclerosis.2 Interleukin-1 (IL-1), tumor necrosis factor-
(TNF-), and
lipopolysaccharides (LPSs) each induce the expression of
endothelial E-selectin and other CAMs, like
intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion
molecule-1 (VCAM-1).1 3 This process can be
inhibited by antioxidants4 5 6 and mimicked by
exogenous hydroperoxides, as has been demonstrated for ICAM-1, though
not for E-selectin.7 It has further been
suggested that reactive oxygen species are somehow involved in the
signaling pathways of all aforementioned inducers of
CAMs.8 These observations led to the intriguing
hypothesis that CAM expression, and therefore the initiation of
atherogenesis, is facilitated by generation of reactive oxygen
species.2 9 The hypothesis is supported by the
fact that the expression of adhesion molecules requires the activation
of nuclear factor-
B (NF-B),10 11 12 which
involves cascades of phosphorylation and
dephosphorylation, but is also modulated by oxidative
events8 13 that can be triggered by
TNF-8 and IL-1.14
However, inhibition of cytokine-induced signals by
antioxidants is not as unequivocal as it may appear at first glance.
The inhibition theory is mainly based on effects observed with
pyrrolidine dithiocarbamate and N-acetylcysteine
(NAC).4 5 6 8 However, pyrrolidine dithiocarbamate
in physiological systems acts as a
pro-oxidant15 16 rather than an antioxidant, and
NAC has also been shown to act as an oxidant under specific
circumstances.17 With this situation in mind, the
inhibitory effects of redox-active compounds can indeed be
based on reductive as well as oxidative processes, depending on the
step in the signaling cascade with which they interfere. This idea is
further supported by some recent publications describing the inhibition
of IL-1– and TNF-–induced NF-B activation by thiol-modifying,
mainly oxidizing, agents,18 19 whereas
thiol-reducing compounds either had no effect or were even
stimulating.18 19 20 Also, LPS-induced lung injury
and adhesion molecule expression were attenuated in rats depleted of
cellular thiols by diethylmaleate.21 Taken
together, the cell type, the particular compound used, the preexisting
redox potential, as well as the thiol status of the cells appear to
influence the effects of redox-active compounds on cytokine
signaling (for a review see Reference 1313 ).
In view of the conflicting results mentioned above and the debatable
benefit of antioxidants in the prevention of atherogenesis, we studied
whether thiol-modifying agents were able to modulate the
cytokine-induced expression of endothelial
E-selectin. These compounds strongly inhibited the IL-1– and
TNF-–induced E-selectin expression, demonstrating that free
(protein) thiols are essential for appropriate
cytokine-mediated signaling.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Cell Culture and Treatments
Human umbilical vein endothelial cells (HUVECs) were obtained by treatment of umbilical veins with 0.2% dispase (Boehringer Mannheim) in Hanks' balanced salt solution (Biochrom). Cells were seeded into gelatin-coated flasks in endothelial cell growth medium containing 2% FBS and 0.4% endothelial cell growth supplement/heparin (EGM; PromoCell). Experiments were performed with confluent cells grown in 48-well Falcon plates between the first and third passage. Stock solutions (1000-fold) of thiol agents were prepared before use as follows: diamide (diazene-dicarboxylic acid bis[N,N'-dimethylamine]), L-buthionine-[S,R]-sulfoximine (BSO), N-ethylmaleimide (NEM), DL-DTT, 2,3-dimercaptopropanol (DMP), and 2-mercaptoethanol (2-ME) were dissolved or diluted in water; diethylmaleate (maleic acid diethyl ester, DEM) was diluted with 1,2-propanediol; phenylarsine oxide (PAO) and monochlorobimane (MBCl; Molecular Probes) were dissolved in dimethylsulfoxide; NAC was dissolved in EGM, and the pH was titrated to 7.4 with 1 mol/L NaOH. All chemicals were purchased from Sigma and were of the highest purity available.
Quantitative Immunofluorescence Labeling
After pretreatment, confluent HUVECs were washed and stimulated
with recombinant human TNF- (Biochrom) or recombinant human IL-1
(kindly provided by the Gesellschaft für Biotechnologische
Forschung, Braunschweig, Germany) at 1 ng/mL EGM each for the times
indicated. After they were washed with PBS, the cells were fixed with
2% formalin in PBS, washed twice for 2 minutes, treated with ice-cold
methanol for 1 minute, and washed with PBS 3 times. E-selectin
expression was analyzed by indirect
immunofluorescence using an anti-human E-selectin
monoclonal mouse antibody (Dako; 1:100 for 1 hour) as the first and a
Cy3 (indocarbocyanin)-conjugated goat anti-mouse IgG (Dianova; 1:100
for 1 hour) as the second antibody. Fluorescence quantification
was performed by scanning in a Cytofluor II microplate
fluorescence reader (Perseptive Biosystems) at 530 nm
excitation and 590 nm emission. Background levels achieved by
incubation with an isotype mouse IgG1 (Sigma,
1:100) as the first antibody were subtracted from the experimental
values obtained.
Total Glutathione and Glutathione Disulfide (GSSG; Kinetic
DTNB Assay)
Total glutathione and glutathione disulfide (oxidized
glutathione) were determined according to Akerboom and
Sies22 adapted to cell cultures. HUVECs grown to
confluence and treated in Petri dishes of 165-mm diameter (Greiner)
were washed with PBS, scraped into ice-cold PBS, and
centrifuged (5 minutes at 600g). For total
glutathione and GSSG determination, cells were resuspended in 600 µL
ice-cold homogenization buffer (50 mmol/L
Tris-HCl, pH 7.4; 1 mmol/L EDTA; and 0.5% wt/vol Triton X-100)
with (for GSSG determination) or without (for total glutathione
determination) 25 mmol/L NEM and sonicated (20 seconds at 25 W).
Aliquots (20 µL) of homogenates were taken for protein
determination23 with the Pierce protein assay
reagent and BSA (both from KMF Laborchemie) as the standard. The
protein of homogenates was precipitated with 1 mol/L (final
concentration) HClO4 for 30 minutes and removed
by centrifugation (10 minutes at 10 000g),
and the supernatant was neutralized with 0.6 mol/L MOPS and 4 mol/L
KOH. Precipitated KClO4 was removed by
centrifugation (10 minutes at 10 000g). NEM
was extracted 4 times with water-saturated ethyl acetate. Residual
solvent was removed with nitrogen. Total glutathione and GSSG were
determined by following the continuous reduction of DTNB by reduced
glutathione (GSH) in the presence of 155 mU/mL glutathione reductase
(Boehringer Mannheim) and 153 µmol/L NADPH (Sigma) at
412 nm. Concentrations were quantified with an internal standard
(0.5 µmol/L GSSG). The concentration of GSH was calculated by
subtracting 2x the GSSG concentration from total glutathione.
Intracellular Thiols
HUVECs were washed with Earle's balanced salt solution, treated
with 100 µmol/L MBCl in Earle's balanced salt solution for 40
minutes at 37°C, washed, and scanned in a Cytofluor II at 390 nm
excitation and 460 nm emission. Background fluorescence of
unlabeled cells was subtracted from the experimental values
obtained.
Cell Viability
Cell viability was assessed by quantification of MTT
(3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide)
reduction by mitochondrial dehydrogenases. In brief, cells were
incubated for 2 hours with 1.2 mmol/L MTT in EGM. After the cells
were washed with Hanks' balanced salt solution, formazan dye was
solubilized in 5% formic acid in isopropanol, and the extinction was
measured at 550 versus 690 nm in a microplate reader.
Protein Synthesis
Protein biosynthesis rates were measured by incorporation of
L-[4,5-3H]leucine (120 to 190
Ci/mmol, Amersham Buchler) into cellular proteins. Confluent cells were
incubated for 2 hours with EGM containing 0.1 mmol/L
L-leucine and 3.5 µCi/mL
L-[4,5-3H]leucine. Cells were
washed with cold PBS, followed by treatment for 30 minutes at 4°C
with 5% trichloroacetic acid. Plates were extensively washed (3x)
with 5% trichloroacetic acid, and the precipitated protein was
solubilized (1 hour) with 1 mol/L NaOH, followed by neutralization with
HCl and scintillation counting (Scintillator Plus, Packard).
Statistical Analysis
Bilateral Student's t test was used to assess
differences at the P<0.01 level.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Induction of E-Selectin Expression by IL-1 and TNF-
IL-1 and TNF-
induced a time-dependent expression of E-selectin
in HUVECs (Figure 1
). E-selectin
expression was maximal 4 hours after cytokine addition. No
basal expression of E-selectin was found. Two hours after
cytokine stimulation, E-selectin production was 
40%
of the maximal level obtained after 4 hours with both cytokines
(Figure 1). The inhibition studies, however, were not performed at the
peaks of E-selectin expression, since inhibitory effects of
the diamide tended to be reversed over time. Therefore, in all further
experiments, cells were stimulated with IL-1 or TNF- for only 2
hours, thereby tolerating a submaximal induction of E-selectin.
|
Inhibition of Cytokine-Induced E-Selectin Expression
by Diamide
Thiol-containing antioxidants such as NAC have been shown to
inhibit cytokine-induced endothelial CAM
expression.4 5 We therefore studied whether NAC
inhibited E-selectin expression in our HUVECs. Surprisingly, cells
pretreated with up to 30 mmol/L NAC for 1 hour, which is suggested
to fully reduce the cellular thiol status, did not show any impaired
cytokine responsiveness (Table 1). Similarly, we also did not observe
any inhibition by NAC of the IL-1–mediated activation of the IL-1
receptor–associated protein kinase in T
lymphocytes.19 In the latter system, in contrast,
thiol-modifying agents proved to be inhibitory. We
therefore tested whether IL-1– and/or TNF-–mediated processes
could be correspondingly influenced in endothelial
cells. We indeed found that pretreatment of HUVECs with 100
µmol/L diamide for 1 hour completely abolished the E-selectin
expression induced by TNF- (Figure 2A, 0 hours) and reduced the IL-1–mediated E-selectin expression to
20% of control values (Figure 2C, 0 hours). When cells were allowed
to recover from diamide pretreatment for 1 to 6 hours, they regained
the ability to respond to cytokines with synthesis of
E-selectin (Figure 2A and 2C, light bars).
|
|
The ability of the cells to recover from diamide pretreatment may be in
line with their capacity to reverse diamide-induced
disturbances of their thiol status by means of glutathione. The
cellular glutathione content was therefore decreased by pretreating the
cells with BSO, a specific inhibitor of
-glutamylcysteine synthetase,24 before
incubation with diamide. Twenty hours after addition of 30
µmol/L BSO, the total glutathione content in HUVECs decreased from
3.5±0.6 nmol/mg protein to about one third that value, ie, 1.2±0.3
nmol/mg protein. As is obvious from Figure 2
(A through D), BSO
pretreatment by itself did not change the response of HUVECs to
cytokines to any significant extent, indicating that
glutathione deficiency alone was not responsible for the inhibition of
E-selectin synthesis by diamide. However, reversal of the depression of
E-selectin expression by 100 µmol/L diamide for 1 hour was
almost abolished by pretreatment with BSO (Figure 2A and 2C, dark
bars). Because the assays showed a distinct reduction in cell viability
after treatment of glutathione-depleted cells with 100 µmol/L
diamide (Table 2), the experiments were
repeated at lower diamide concentration. Diamide at 30 µmol/L
inhibited the TNF-–induced E-selectin expression by 50% (Figure 2B, 0 hours, light bars). Full recovery was already observed after 2
hours. The IL-1 effect was not influenced by 30 µmol/L diamide
(Figure 2D, light bars). In glutathione-deficient cells, however,
30 µmol/L diamide completely depressed the TNF-– as well as
the IL-1–induced E-selectin expression, with only marginal effects on
cell viability (Table 2). Under these conditions, the response to
cytokines was not restored immediately but only after a 6-hour
recovery period in growth medium (Figure 2B and 2D, dark bars). This
result shows that the ability of the cells to restore the TNF- and
IL-1 signaling cascade after diamide treatment is not absolutely
dependent on optimally filled glutathione pools. However, high
glutathione levels strongly facilitate recovery.
|
Cellular Thiol Status After Diamide Treatment
Glutathione
To further elucidate the role of glutathione in
cytokine-induced E-selectin expression, we measured
cellular glutathione status after diamide treatment of
HUVECs. Owing to the limitation of sample size, direct determination of
GSH by, eg, GSH S-transferase,25 was
not feasible. Therefore, we used the kinetic DTNB
assay22 for estimating the concentration of total
(reduced plus oxidized) glutathione as well as of GSSG. By this assay,
a total glutathione content of 3.0±0.7 nmol/mg protein was assessed in
our cells, a value that seems low compared with previously published
data.26 However, it is not the absolute
glutathione concentration but the ratio of GSH to GSSG that
characterizes the cellular redox status.25 This
ratio was 27 in our cells, which is high when compared with those
published elsewhere (11 after transformation into molar
levels).26
Total glutathione in HUVECs did not change to any significant extent
after treatment with 100 µmol/L diamide for 1 hour or during the
following recovery period of up to 4 hours (Table 3). Because of the relatively high SD of
the cellular glutathione concentrations (Table 3), a putatively small
decrease in GSH after diamide treatment was not detectable. However,
GSH tended to increase during the recovery period, reaching
significance 4 hours after diamide treatment (Table 3).
|
As anticipated, GSSG increased from 0.1 to 0.29 nmol/mg protein after
diamide treatment (Table 3). This resulted in a distinct shift in the
GSH to GSSG ratio, from 27.5 in control cells to 9.2 in diamide-treated
cells, which returned to control values during 1 hour of recovery
(Table 3). Thus, normalization of glutathione status preceded the
recovery of cytokine-induced E-selectin expression, which
was restored only after 4 to 6 hours.
Glutathione and Protein Thiols
To detect diamide-induced changes in the overall thiol status of
individual cell populations, we used the MBCl labeling
technique.27 MBCl, by itself
nonfluorescent, forms stable, fluorescent adducts with
thiols. It has been shown to specifically react with GSH due to
catalysis by GSH S-transferases in rodent cell lines, but
not in human cells.28 Therefore, MBCl labeling of
HUVECs is supposed to result predominantly from the (slower)
nonenzymatic reaction of MBCl with all kinds of thiols, including
protein thiols.
By MBCl labeling, thiols decreased by 20% in HUVECs treated with
100 µmol/L diamide and returned to control values after 2 hours
(Figure 3A). This time course was similar
to that of E-selectin recovery after diamide treatment (Figure 2A and 2C). In glutathione-deficient cells, the thiol level was 40% of
controls and was decreased further by 30 µmol/L diamide (Figure 3B). During subsequent incubation in EGM for up to 6 hours, the thiol
content recovered but did not reach the control level, ie, that of
BSO-treated cells (Figure 3B). In contrast, the inhibition of
E-selectin expression was almost completely reversed during the same
time interval (Figure 2B and 2D). The data clearly indicate that
impairment and restoration of cytokine signaling do not
parallel the shifts in the total thiol status or in the GSH to GSSG
ratio. More likely, it depends on specific thiols, in particular
proteins, which are prone to oxidation but are regenerated
preferentially despite persisting disturbance of thiol status
in GSH-depleted cells.
|
Inhibition of Cytokine-Induced E-Selectin Expression
by PAO
To gain further insights into the nature of the protein thiols
involved in IL-1– and TNF-–mediated E-selectin expression, we
analyzed the effect of PAO, an agent that has been described to
react with vicinal protein thiols.29
Cytokine-induced E-selectin expression was completely
depressed by pretreatment of cells with 0.5 µmol/L PAO for 20
minutes (Figure 4A), an effect that was
almost completely reversed by the dithiols DTT and DMP but not by the
monothiol 2-ME (Figure 4A). In contrast to diamide-treated cells,
PAO-treated HUVECs did not recover spontaneously in culture medium
within 4 hours (not shown).
|
As shown by MBCl labeling (Figure 4B), PAO induced a 20% reduction in
cellular thiols in HUVECs. This diminishment corresponds to an apparent
decrease in cell viability as measured by MTT reduction (Figure 4B).
The latter, however, resulted from a loss of cells due to detachment
during the washing processes. Because the remaining cells did not
change their ability to reduce MTT during the whole experimental
period, we can state that the viability and thiol content of HUVECs
were not significantly influenced by PAO treatment. We can further
state that the restoration by DTT and DMP of E-selectin expression
(Figure 4A) to 80% corresponds to 100% of the remaining cells.
Total thiol levels in PAO-treated cells remained unaffected by DTT or
DMP (Figure 4B), although these dithiols completely reversed the
inhibitory effect of PAO on cytokine-induced
E-selectin expression (Figure 4A). This finding again underlines the
concept that the thiol status of specific proteins is necessary for
cytokine signaling.
Effects of PAO on Protein Synthesis Rates
Because it has been shown that PAO inhibits protein
biosynthesis,30 we tested whether this could be
the reason for the observed inhibition of E-selectin expression. PAO at
0.5 µmol/L inhibited E-selectin expression by nearly 100%,
whereas protein biosynthesis at the same time was also depressed but
still active, at 1/3 of its original activity (Figure 5). Protein synthesis rates indeed
increased during the recovery period in EGM and EGM containing 2-ME,
which, however, had no effect on E-selectin expression. Only on
complete reversal of PAO by DTT or DMP did all parameter
return to control levels (Figure 5). Thus, inhibition of the protein
biosynthesis rate by PAO can only be part of the PAO-induced inhibition
of E-selectin expression. In addition, diamide did not influence the
protein biosynthesis rate by >20% under all conditions investigated,
except after 100 µmol/L diamide in BSO-treated cells. Here
again, the protein synthesis rate was 1/3 that of control (not shown),
but E-selectin expression was depressed by 100%.
|
Inhibition of E-Selectin Expression by Other Thiol-Modifying
Compounds
In addition to diamide and PAO, we tested other compounds known to
modify sulfhydryl groups. The thiol-alkylating agents
NEM31 and DEM32 both
inhibited TNF-– and IL-1–mediated E-selectin expression in a
concentration-dependent manner (Table 1) but did not influence cell
viability (Table 2) or protein synthesis (not shown) at all. A 100%
inhibition of TNF-–mediated E-selectin expression was obtained with
1 mmol/L DEM for 1 hour and with 5 µmol/L NEM for 30
minutes. IL-1–mediated E-selectin expression was influenced similarly,
except for NEM, which led to only a 58% inhibition at a concentration
of 5 µmol/L for 30 minutes.
Cellular Viability After Treatment With Thiol-Modifying
Agents
DEM and NEM tested at the same concentrations to inhibit
E-selectin expression (1 mmol/L and 5 µmol/L, respectively)
did not influence cell viability at 0, 4, and 24 hours after treatment,
as measured by MTT assays (Table 2). Diamide (30 µmol/L)
slightly decreased cell viability in glutathione-depleted cells at time
0 but not at 4 or 24 hours later (Table 2). PAO at 0.5 µmol/L
(20 minutes) or diamide at 100 µmol/L (1 hour) reduced viability
by 15% to 20%, which, however, was mainly caused by detachment and
loss of cells during the washing processes, as mentioned before. Only
PAO produced an additional decrease in cell viability after 24 hours
(Table 2), which, however, was not a time point relevant for the
present study.
Incubation of BSO-treated cells with 100 µmol/L diamide
decreased cell viability during the whole experimental period to
60% to 70% of controls (Table 2). However, this decrease in
viability by 40% was not correlated with the complete depression of
E-selectin expression up to 6 hours after treatment with the high dose
of diamide.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Herein we have shown inhibition of cytokine-mediated E-selectin expression in HUVECs by thiol-modifying agents, whereas NAC (up to 30 mmol/L) was without any effect. Thus, a direct antioxidant or reductive effect of NAC seems to play a subordinate role in our HUVEC system, if any. NAC effects are seemingly highly controversial. In addition, the effects of NAC on signaling processes differ in various cell lines. For example, Brennan and O'Neill20 demonstrated NAC to inhibit NF-
B
activation in cytokine-induced Jurkat T cells but not in
EL4.NOB-1 and KB cells.20 Similarly, we could not
confirm any effect of NAC on IL-1–mediated NF-B activation in the
murine EL4 6.1 cell line.19 Also, no influence of
NAC on IL-1–mediated activation of an IL-1 receptor–associated
protein kinase was found in T cells19 as well as
ECs.33 Consistently, however, we observed
inhibition of the activation of this kinase by a series of
thiol-modifying agents.19 33 This finding is in
accord with the results of the present study, ie, the inhibition of
IL-1– as well as TNF-–induced E-selectin expression by
preincubation with subtoxic concentrations of several thiol-modifying
agents (viz, diamide, PAO, NEM, and DEM). All of these compounds have
in common the propensity to react with low- and high-molecular-weight
thiols.
Diamide has been described to rapidly react with GSH, to yield GSSG and
a hydrazine derivative.34 Protein thiols,
including some with vicinal thiols like thioredoxin, react with diamide
at distinct, usually lower reaction rates, yielding protein disulfides
or mixed disulfides.35 Incubation of HUVECs with
100 µmol/L diamide strongly depressed TNF-– and
IL-1–induced E-selectin expression. At the same time, the
concentration of GSSG was increased by a factor of 3. During the
recovery period, the cells regained their ability to respond to
cytokines with E-selectin expression, unless they were made
glutathione deficient by BSO. These data show that modification of free
thiols strongly impairs cytokine-induced expression of
E-selectin and that intracellular GSH may counteract this inhibition.
The diamide-induced disturbance in glutathione status per se
cannot, however, explain the inhibition of E-selectin expression for
the following reasons: (1) the enhancement of GSSG by 100 µmol/L
diamide was normalized after 1 hour, but E-selectin expression remained
depressed and (2) decreasing the glutathione content with BSO by itself
did not influence cytokine-induced E-selectin expression.
Also, diamide does not affect the level of reduced glutathione. Its
effect on the glutathione system is detectable only as a transient
increase in GSSG (Table 3). In contrast, MBCl fluorescence,
which represents all kinds of thiols, drops significantly on
treatment with diamide (Figure 3). Therefore, we must conclude that in
our system, diamide preferentially reacts with particular protein
thiols, although all cellular thiol pools are intimately interrelated.
In control cells, protein thiols consistently recover within 4
to 6 hours, whereas the thiol content in BSO-treated cells stays
significantly below starting conditions for 6 hours after diamide
treatment (Figure 3). Thus, the protein thiol/disulfide equilibrium
indeed depends on glutathione, but quantitatively, the
disturbance in the protein thiol/disulfide status on treatment
with diamide is more severe than that of the glutathione system.
Therefore, it is not unreasonable to presume a modification of some
protein thiols to be responsible for the observed inhibition of
E-selectin expression by diamide.
This notion is further supported by the results obtained with the other thiol-modifying agents. DEM and NEM are cell-permeant, thiol-alkylating agents that were first used to deplete cells of glutathione.31 32 Both, however, rapidly react with protein thiols, and NEM is now used to block all accessible cellular protein sulfhydryls, thereby "freezing" the actual redox state of cells for studies of redox regulation.36
The most intriguing information on the involvement of critical protein
sulfhydryls in the TNF- and IL-1 signaling in HUVECs comes from the
experiments with PAO. PAO forms dithioarsine adducts of high stability,
with 2 suitably spaced (mostly vicinal) sulfhydryl groups of proteins.
Vicinal thiols are often exposed at the protein surface, thus allowing
interactions with substrates, oxidants, and other proteins that
catalyze disulfide reduction.37 Enzymes
containing vicinal thiols in the active center include
ribonucleotide reductase, glutathione reductase,
thioredoxin reductase, thioredoxin, and glutaredoxin (for a review, see
Reference 3737 ). PAO has been demonstrated to inhibit insulin functions,
receptor internalization, and platelet
activation.37 In addition, specific enzymes are
affected, eg, phosphotyrosine phosphatases.38 We
have shown herein that PAO at concentrations as low as 0.5
µmol/L fully depresses TNF-– and IL-1–mediated E-selectin
expression. This inhibition can be reversed by short-term treatment of
the cells with dithiols like DMP or DTT but not with the monothiol
2-ME, which is characteristic for reversal of PAO binding to vicinal
protein thiols.29 The inhibition of E-selectin
expression by PAO may result to some extent from general inhibition of
protein biosynthesis to 1/3 of control values. The latter process,
however, cannot explain the total inability of the cells to respond to
cytokines, nor can it explain the much less pronounced effect
of diamide on protein biosynthesis, which was able to block E-selectin
expression. Thus, in addition to the proteins involved in protein
biosynthesis and other physiological processes, the
proteins crucial for IL-1 and TNF- signaling most probably contain
vicinal thiols.
The nature of the affected proteins and the site and time of their
function in the signaling cascade are not easy to determine from our
results. Several considerations must be taken into account: (1) We
recently demonstrated inhibition of the activation of an IL-1 receptor
type I (IL-1RI)–associated protein kinase, one of the earliest events
in IL-1 signaling,39 by the very same
thiol-modifying agents used in this study.19 The
kinase was first described and has mainly been investigated in T
lymphocytes, but it is also present in an EC
line.33 It may thus be concluded that activation
of the IL-1RI–associated kinase finally leads to E-selectin expression
in ECs. Association of the kinase with the IL-1 receptor makes it
specific for IL-1–mediated signals. It remains possible, however, that
receptor-associated kinases are also involved in TNF- signaling and
that thiol modification thereof has similar effects as it does on the
IL-1 receptor–associated kinase. (2) A common signal mediated by both
cytokines is the activation of NF-B. NF-B is necessary,
though not sufficient, for the induction of
E-selectin.10 11 Efficient binding of NF-B to
its responsive elements and thus, unimpaired
trans-activating activity depend on a cysteine thiol in the
p50 subunit of the p50/p65 heterodimer being kept in a reduced state by
thioredoxin.40 41 If thioredoxin is
inactivated by oxidative processes, general inhibition of
gene activation may be the consequence, since thioredoxin has also been
shown to regulate via redox factor 1 the activity of activating
transcription factor,42 43 another transcription
factor involved in E-selectin gene
transcription.10 (3) Both IL-1 and TNF- have
been shown to be internalized by a receptor-mediated
process.44 45 Suppression of TNF-–mediated
E-selectin expression by PAO has been attributed to an inhibition of
receptor-mediated endocytosis of TNF-.30
Whether this mechanism can explain the observations with IL-1 remains
to be established. (4) Protein tyrosine phosphorylation
plays a major role in the regulation of a number of cellular
metabolic pathways, among them cellular
adhesion.46 It is regulated by the opposing
actions of protein tyrosine kinases and protein tyrosine phosphatases
(PTPs) and therefore requires a well-balanced activity of both enzymes.
Both kinases and phosphatases have been shown to be targets for redox
regulation (reviewed in Reference 4747 ) and are therefore highly
sensitive to disturbances in the cellular redox balance. In
addition, all PTPs have a conserved cysteine residue in their catalytic
domain which must be in the reduced form for full
activity.47 Inhibition of PTPs, however, results
in unopposed kinase action and, in consequence, an amplified
signal.38 This is in line with the observation
that inactivation of a PTP by sulfhydryl-modifying compounds mimics
early effects of TNF- and IL-1 in human
fibroblasts.48 Since we observed an inhibition of
cytokine-mediated E-selectin expression and not an
amplification thereof, it may be doubted whether PTP inactivation plays
a major role in our system. This might also be considered in the
interpretation of the recently described inhibition of
endothelial CAM expression by diamide and PAO, which
was discussed, however not shown, to be a PTP inactivation
process.49
In essence, our data show that E-selectin expression requires proteins
with free, most probably vicinal, thiols for adequate inducibility by
IL-1 or TNF-. The reduced state of crucial proteins may be
maintained by glutathione, the major cellular redox buffer, since
depletion of GSH dramatically enhances the sensitivity of target
proteins for modifications. However, the time course of recovery after
diamide treatment of GSH compared with E-selectin does not convincingly
support an obligatory reduction of oxidized protein thiols by
glutathione. It seems more likely that glutathione provides the
reducing equivalents for a second system. Vicinal thiols are kept in
the reduced state either by the thioredoxin or the glutaredoxin
system.50 Which of the systems plays the main
role in our model remains to be investigated.
Our data are not meant to rule out an induction of adhesion molecules
by physiological oxidants such as
H2O2 or
lipoxygenase products. In fact, the evidence for an
induction of leukocyte sticking51 and of the
expression of adhesion molecules by eg, oxidized low density
lipoprotein, is overwhelming2 9 and with good
reason is considered an initial event in vascular inflammation,
potentially leading ia to atherogenesis. Rather, our findings indicate
that some component in the IL-1 and/or TNF- signaling cascade may
become refractory52 when its thiols are oxidized
or permanently modified. The latter, however, under specific
circumstances, may be beneficial for an organism, as was demonstrated
for the inhibition of LPS-induced rat lung injury by DEM
pretreatment.21 Thus, a reduced status of the
signaling cascade responding to IL-1 and TNF- in ECs is a
prerequisite for a normal immune response, which apparently comprises
the oxidation of vicinal thiols in a signaling protein, whereas
previous oxidation or modification of these thiols renders the system
unresponsive.
|
Acknowledgments |
|---|
This study was supported by the Deutsche Forschungsgemeinschaft (to Dr Friedrichs; FR 1160/1–1). The excellent technical assistance of Hella Blumhagen and Ilse Wölfel is gratefully acknowledged.
Received December 4, 1997; accepted May 8, 1998.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Bevilacqua MP, Pober JS, Mendrick SL, Cotran RS, Gimbrone MA Jr. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci U S A. 1987;84:9238–9242.
2. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature. 1993;362:801–809.[Medline] [Order article via Infotrieve]
3. Beekhuizen H, van Furth R. Monocyte adherence to human vascular endothelium. J Leukoc Biol. 1993;54:363–378.[Abstract]
4.
Faruqi RM, Poptic EJ, Ruhe Faruqi T, De La Motte C,
DiCorleto PE. Distinct mechanisms for N-acetylcysteine inhibition of
cytokine-induced E-selectin and VCAM-1 expression.
Am J Physiol. 1997;273:H817–H826.
5. Marui N, Offerman M, Swerlick R, Kunsch C, Rosen C, Ahmad M, Alexander W, Medford R. Vascular cell adhesion molecule-1 (VCAM-1) gene transcription and expression are regulated through an antioxidant-sensitive mechanism in human vascular endothelial cells. J Clin Invest. 1993;92:1866–1874.
6. Kawai M, Nishikomori R, Jung EY, Tai G, Yamanaka C, Mayumi M, Heike T. Pyrrolidine dithiocarbamate inhibits intercellular adhesion molecule-1 biosynthesis induced by cytokines in human fibroblasts. J Immunol. 1995;154:2333–2341.[Abstract]
7. Bradley JR, Johnson DR, Pober JS. Endothelial activation by hydrogen peroxide: selective increases of intercellular adhesion molecule-1 and major histocompatibility complex class I. Am J Pathol. 1993;142:1598–1609.[Abstract]
8.
Schreck R, Rieber P, Baeuerle PA. Reactive oxygen
intermediates as apparently widely used messengers in the activation of
the NF-B transcription factor and HIV-1. EMBO J. 1991;10:2247–2258.[Medline]
[Order article via Infotrieve]
9. Witztum JL. The oxidation hypothesis of atherosclerosis. Lancet. 1994;344:793–795.[Medline] [Order article via Infotrieve]
10.
De Luca LG, Johnson DR, Whitley MZ, Collins T, Pober
JS. cAMP and tumor necrosis factor competitively regulate
transcriptional activation through and nuclear factor binding to the
cAMP-responsive element/activating transcription factor element of the
endothelial leukocyte adhesion molecule-1 (E-selectin)
promoter. J Biol Chem. 1994;269:19193–19196.
11.
Lewis H, Kaszubska W, DeLamarter JF, Whelan J.
Cooperativity between two NF-B complexes, mediated by
high-mobility-group protein I(Y), is essential for
cytokine-induced expression of the E-selectin promoter.
Mol Cell Biol. 1994;14:5701–5709.
12.
Collins T, Read MA, Neish AS, Whitley MZ, Thanos D,
Maniatis T. Transcriptional regulation of endothelial
cell adhesion molecules: NF-B and cytokine-inducible
enhancers. FASEB J. 1995;9:899–909.[Abstract]
13.
Flohé L, Brigelius-Flohé R, Saliou C,
Traber MG, Packer L. Redox regulation of NF-B activation.
Free Radic Biol Med. 1997;22:1115–1126.[Medline]
[Order article via Infotrieve]
14.
Brigelius-Flohé R, Friedrichs B, Maurer S,
Schultz M, Streicher R. Interleukin-1-induced nuclear factor B
activation is inhibited by overexpression of phospholipid hydroperoxide
glutathione peroxidase in a human endothelial cell
line. Biochem J. 1997;328:199–203.
15.
Nobel CSI, Kimland M, Lind B, Orrenius S, Slater AFG.
Dithiocarbamates induce apoptosis in thymocytes by raising the
intracellular level of redox-active copper. J Biol
Chem. 1995;270:26202–26208.
16.
Galter D, Mihm S, Dröge W. Distinct effects of
glutathione disulfide on the nuclear transcription factors B
and the activator protein-1. Eur J Biochem. 1994;221:639–648.[Medline]
[Order article via Infotrieve]
17.
Das KC, Lewis-Molock Y, White CW. Activation of
NF-B and elevation of MnSOD gene expression by thiol reducing
agents in lung adenocarcinoma (A549) cells. Am J
Physiol. 1995;269:L588–L602.
18.
Das KC, Lewis-Molock Y, White CW. Thiol modulation of
TNF and IL-1 induced MnSOD gene expression and activation of
NFB. Mol Cell Biochem. 1995;48:45–57.
19.
Tewes F, Böl G-F, Brigelius-Flohé R. Thiol
modulation inhibits the interleukin (IL)-1-mediated activation of an
IL-1 receptor-associated protein kinase and NFB. Eur
J Immunol. 1997;27:3015–3021.[Medline]
[Order article via Infotrieve]
20.
Brennan P, O'Neill LAJ. Effects of oxidants and
antioxidants on nuclear factor B activation in three different
cell lines: evidence against a universal hypothesis involving oxygen
radicals. Biochim Biophys Acta. 1995;1260:167–175.[Medline]
[Order article via Infotrieve]
21. Nathens AB, Marshal JC, Watson RWG, Dackiw APB, Rotstein OD. Diethylmaleate attenuates endotoxin-induced lung injury. Surgery. 1996;120:360–366.[Medline] [Order article via Infotrieve]
22. Akerboom TP, Sies H. Assay of glutathione, glutathione disulfide, and glutathione mixed disulfides in biological samples. Methods Enzymol. 1981;77:373–382.[Medline] [Order article via Infotrieve]
23. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principles of protein-dye binding. Anal Biochem. 1976;72:248–254.[Medline] [Order article via Infotrieve]
24. Meister A. Glutathione biosynthesis and its inhibition. Methods Enzymol. 1995;252:26–30.[Medline] [Order article via Infotrieve]
25. Brigelius R, Muckel C, Akerboom TPM, Sies H. Identification and quantification of glutathione in hepatic protein mixed disulfides and its relationship to glutathione disulfide. Biochem Pharmacol. 1983;32:2529–2534.[Medline] [Order article via Infotrieve]
26.
Jornot L, Junod AF. Variable glutathione levels and
expression of antioxidant enzymes in human endothelial
cells. Am J Physiol. 1993;264:L482–L489.
27. Mirabelli F, Vairetti M, Bellomo G. Fluorescence analysis of intracellular soluble and protein thiols using bimane derivatives. Appl Fluorescence Technol. 1992;4:9–13.
28.
Uplacker GA, Johnson JA, Siegel FL, Mulcahy RT.
Influence of glutathione S-transferases on cellular glutathione
determination by flow cytometry using monochlorobimane. Cancer
Res. 1991;51:1783–1788.
29. Kalef E, Walfish PG, Gitler C. Arsenical-based affinity chromatography of vicinal dithiol-containing proteins: purification of L1210 leukemia cytoplasmic proteins and the recombinant rat c-erb Aß1 T3 receptor. Anal Biochem. 1993;212:325–334.[Medline] [Order article via Infotrieve]
30. Bradley JR, Johnson DR, Pober JS. Four different classes of inhibitors of receptor-mediated endocytosis decrease tumor necrosis factor-induced gene expression in human endothelial cells. J Immunol. 1993;150:5544–5555.[Abstract]
31. Beutler E, Srivastava SK, West C. The reversibility of N-ethylmaleimide (NEM) alkylation of red cell glutathione. Biochem Biophys Res Commun. 1970;38:341–347.[Medline] [Order article via Infotrieve]
32. Thor J, Moldeus P, Hermanson R, Hoegberg J, Reed DJ, Orrenius S. Metabolic activation and hepatotoxicity: toxicity of bromobenzene in hepatocytes isolated from phenobarbital- and diethylmaleate-treated rats. Arch Biochem Biophys. 1978;188:122–129.[Medline] [Order article via Infotrieve]
33. Brigelius-Flohé R, Friedrichs B, Tewes F, Böl G-F. Redox events in the cytokine-induced expression of adhesion molecules in endothelial cells. 1998 World Congress of the Oxygen Club of California, `Oxidants and Antioxidants in Biology,' February 5–8, 1998, Santa Barbara, California. Abstract.
34. Kosower NS, Kosower EM, Wertheim B, Correa WS. Diamide, a new reagent for the intracellular oxidation of glutathione to the disulfide. Biochem Biophys Res Commun. 1969;37:593–596.[Medline] [Order article via Infotrieve]
35. Kosower NS, Kosower EM. Diamide: an oxidant probe for thiols. Methods Enzymol. 1995;251:123–133.[Medline] [Order article via Infotrieve]
36. Gitler C, Zarmi B, Kalef E. General method to identify and enrich vicinal thiol proteins present in intact cells in the oxidized, disulfide state. Anal Biochem. 1997;252:48–55.[Medline] [Order article via Infotrieve]
37. Gitler C, Mogyoros M, Kalef E. Labeling of vicinal dithiols: role of protein-S2 to protein-(SH)2 conversion in metabolic regulation and oxidative stress. Methods Enzymol. 1994;233:403–415.[Medline] [Order article via Infotrieve]
38.
Garcia-Morales P, Minami Y, Luong E, Klausner RD,
Samelson LE. Tyrosine phosphorylation in T cells is
regulated by phosphatase activity: studies with phenylarsine oxide.
Proc Natl Acad Sci U S A. 1990;87:9255–9259.
39. Martin M, Böl GF, Eriksson A, Resch K, Brigelius-Flohé R. Interleukin-1-induced activation of a protein kinase co-precipitating with the type I interleukin-1 receptor in T cells. Eur J Immunol. 1994;24:1566–1571.[Medline] [Order article via Infotrieve]
40.
Matthews JR, Wakasugi N, Virelizier JL, Yodoi J, Hay
RT. Thioredoxin regulates the DNA binding activity of NF-B by
reduction of a disulphide bond involving cysteine 62. Nucleic
Acids Res. 1992;20:3821–3830.
41.
Hayashi T, Ueno Y, Okamoto T. Oxidoreductive regulation
of nuclear factor B: involvement of a cellular reducing
catalyst thioredoxin. J Biol Chem. 1993;268:11380–11388.
42. Xanthoudakis S, Miao G, Wang F, Pan YCE, Curran T. Redox regulation of fos-jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J. 1992;11:3323–3335.[Medline] [Order article via Infotrieve]
43.
Hirota K, Matsui M, Iwata S, Nishiyama A, Mori K, Yodoi
J. AP-1 transcriptional activity is regulated by a direct association
between thioredoxin and Ref-1. Proc Natl Acad Sci U S A. 1997;94:3633–3638.
44.
Tsujimoto M, Yip YK, Vilcek J. Tumor necrosis factor:
specific binding and internalization in sensitive and resistant
cells. Proc Natl Acad Sci U S A. 1985;82:7626–7630.
45.
Grenfell S, Smithers N, Miller K, Solari R.
Receptor-mediated endocytosis and nuclear transport of human
interleukin 1. Biochem J. 1989;264:813–822.[Medline]
[Order article via Infotrieve]
46.
Clark EA, Brugge JS. Integrins and signal transduction
pathways: the road taken. Science. 1995;268:233–239.
47. Monteiro HP, Stern S. Redox modulation of tyrosine phosphorylation-dependent signal transduction pathways. Free Radic Biol Med. 1996;21:323–333.[Medline] [Order article via Infotrieve]
48.
Guy GR, Cairns J, Ng SB, Tan YH. Inactivation of a
redox-sensitive protein phosphatase during the early events of tumor
necrosis factor/interleukin-1 signal transduction. J Biol
Chem. 1993;268:2141–2148.
49.
Dhawan S, Singh S, Aggarwal BB. Induction of
endothelial cell surface adhesion molecules by tumor
necrosis factor is blocked by protein tyrosine phosphatase
inhibitors: role of the nuclear transcription factor
NF-B. Eur J Immunol. 1997;27:2172–2179.[Medline]
[Order article via Infotrieve]
50.
Holmgren A. Thioredoxin and glutaredoxin systems.
J Biol Chem. 1989;264:13963–13966.
51. Del Maestro R, Thaw HH, Bjork J, Planker M, Arfors KE. Free radicals as mediators of tissue injury. Acta Physiol Scand Suppl. 1980;482:43–57.
52.
Karmann K, Min W, Fanslow WC, Pober JS. Activation and
homologous desensitization of human endothelial cells
by CD40 ligand, tumor necrosis factor, and interleukin-1. J Exp
Med. 1996;184:173–182.
This article has been cited by other articles:
 |
|
 |
|
  N. Jurrmann, R. Brigelius-Flohe, and G.-F. Bol Curcumin Blocks Interleukin-1 (IL-1) Signaling by Inhibiting the Recruitment of the IL-1 Receptor-Associated Kinase IRAK in Murine Thymoma EL-4 Cells J. Nutr., August 1, 2005; 135(8): 1859 - 1864. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-J. ZHANG and B. FREI {alpha}-Lipoic acid inhibits TNF-{alpha}-induced NF-{kappa}B activation and adhesion molecule expression in human aortic endothelial cells FASEB J, November 1, 2001; 15(13): 2423 - 2432. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. ZAHLER, C. KUPATT, and B. F. BECKER Endothelial preconditioning by transient oxidative stress reduces inflammatory responses of cultured endothelial cells to TNF-{alpha} FASEB J, March 1, 2000; 14(3): 555 - 564. [Abstract] [Full Text]
|
|
|
|
 |
|
 P. Wiesel, L. C. Foster, A. Pellacani, M. D. Layne, C.-M. Hsieh, G. S. Huggins, P. Strauss, S.-F. Yet, and M. A. Perrella Thioredoxin Facilitates the Induction of Heme Oxygenase-1 in Response to Inflammatory Mediators J. Biol. Chem., August 4, 2000; 275(32): 24840 - 24846. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||