(Received for publication, September 16, 1996, and in revised form, December 31, 1996)
From the Department of Medicine, Ohio State University, Columbus, Ohio 43210
Interleukin-1 (IL-1
) significantly
influences renal cellular function through the induction of several
gene products. The molecular mechanisms involved in gene regulation by
IL-1
are poorly understood; however, the appearance of novel
tyrosine phosphoproteins in IL-1
-treated cells suggests that IL-1
may function through tyrosine phosphoprotein intermediates. The
mitogen-activated protein (MAP) kinases are tyrosine phosphoproteins
that could potentially mediate the effects of IL-1
. Protein tyrosine
phosphorylation following IL-1
treatment may be dependent on redox
changes since the IL-1
receptor is not a protein-tyrosine kinase and
oxidation has been shown to induce tyrosine phosphorylation. In this
report we demonstrate that conditioning human glomerular mesangial
cells with IL-1
results in the tyrosine phosphorylation and
activation of two members of the MAP kinase family, extracellular
signal-regulated protein kinase 2 (ERK2) and p54
Jun-NH2-terminal kinase (JNK). This effect of IL-1
is abrogated by pretreating cells with the antioxidants
N-acetyl-L-cysteine or dithiothreitol.
Furthermore, the effects of IL-1
on ERK and JNK activation are
reproduced by treating mesangial cells with membrane-permeable
oxidants. IL-1
and oxidants also cause phosphorylation and
activation of the upstream ERK regulatory element MAP kinase kinase.
Interestingly, IL-1
, but not exogenous oxidants, causes
phosphorylation of the upstream JNK activator, JNK kinase. These data
indicate that IL-1
activates ERK2 through an
oxidation-dependent pathway. Exogenous oxidants and IL-1
activate JNK through different upstream mechanisms; however,
antioxidant inhibition of JNK activation indicates that endogenous
oxidants may play a role in IL-1
-induced JNK activation. Thus
IL-1
may affect mesangial cell function by activating MAP kinases,
which can then regulate gene transcription. Furthermore, reactive
oxygen species released during inflammatory glomerular injury may also
affect mesangial function through a MAP kinase signal.
Interleukin-1 (IL-1
)1 has
received considerable attention as a potential mediator of renal injury
(reviewed in Refs. 1-3). IL-1
affects proliferation of cultured
renal cells (4-6) and induces renal cell production of a variety of
biologic modifiers (4, 7-10). IL-1
has been found in the kidneys of
patients and experimental animals with glomerulonephritis (11-15), and
blocking the IL-1
receptor was recently shown to attenuate injury in
experimental glomerulonephritis (16, 17). Thus, IL-1
may be
relevant to the pathogenesis of glomerular damage in
vivo.
Despite a growing understanding of the effects of IL-1 on the
kidney, the intracellular mechanisms that mediate the actions of
IL-1
are poorly characterized. It is known, however, that IL-1
treatment of mesangial cells up-regulates tyrosine kinase activity and
causes tyrosine phosphorylation of several mesangial cell proteins (10,
18-20). Therefore, IL-1
may activate gene transcription in human
mesangial cells through the induction of tyrosine phosphoprotein
intermediates.
In considering potential tyrosine phosphoprotein mediators of IL-1
activity, MAP kinases are of particular interest for several reasons.
These kinases require tyrosine phosphorylation for activity, are
cytokine-inducible, and have been implicated in regulating gene
transcription (21-24). Furthermore, members of the ERK and JNK
subgroups of the MAP kinase family have molecular masses similar in
size (24, 25) to the tyrosine phosphoproteins that are induced by
IL-1
treatment of human mesangial cells (18, 19).
Although IL-1 can induce protein tyrosine phosphorylation, its
receptor is not a classic receptor tyrosine kinase (26). IL-1
-induced tyrosine phosphorylation may be explained by the observation that protein tyrosine phosphorylation can be modulated by
intracellular redox potential. Increasing intracellular oxidation can
directly enhance protein tyrosine phosphorylation (27-29) and can also
potentiate protein tyrosine phosphorylation induced by other stimuli
(30, 31). Furthermore, cytokines such as IL-1
have been reported to
increase oxygen radical release by mesangial cells (32).
This investigation was thus undertaken to determine whether
IL-1-induced tyrosine phosphorylation in human mesangial cells activates ERK and/or JNK and to establish the role of intracellular oxidation in this pathway.
Human mesangial cells were cultured from kidneys not suitable for transplantation and characterized as we have previously described (8). In this investigation, cells from four different donors were used between passages 5 and 10. Cells were grown to confluence in RPMI 1640 containing 10% fetal bovine serum (Life Technologies, Inc.). Before conditioning, monolayers were washed to remove serum; all subsequent incubations were carried out in medium consisting of RPMI 1640 plus 0.25% bovine serum albumin.
In these experiments, mesangial cells were treated with medium alone
(control) or with 1.1 ng/ml human recombinant IL-1 (specific activity, 5 × 108 units/mg; Genzyme Corp., Cambridge,
MA). This concentration of IL-1
was previously determined to cause a
significant induction of tyrosine phosphoproteins in mesangial cells
(18). The redox status of the cells was manipulated by treatment with
the antioxidants DTT or NAC (Sigma) or the cell-permeable oxidizing
agents diamide (33) or hydrogen peroxide (Sigma). In some experiments,
mesangial cells were conditioned with PD 098059 (34, 35), a specific inhibitor of MEK1 and MEK2 activation (a gift from Dr. Alan Saltiel, Parke-Davis, Ann Arbor, MI). All reagents were diluted in RPMI 1640 plus 0.25% bovine serum albumin to achieve the indicated concentrations.
Tissue culture medium and cytokines were found to be virtually endotoxin free (<0.1 ng/ml) using the quantitative Limulus amebocyte lysate assay (M. A. Bioproducts, Walkersville, MD).
Immunoblotting and Immunoprecipitation of Mesangial Cell LysatesConfluent mesangial cells grown in 100-mm plates were appropriately conditioned and, at specified times after treatment, lysed in buffer containing 1% Nonidet P-40, 20 mM Tris-HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 1 mM Na3VO4, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 20 µM leupeptin, 0.15 unit/ml aprotinin. For immunoblotting, 50-100 µg of cell protein were separated on a 10% SDS-polyacrylamide gel under reducing conditions unless otherwise indicated and then transferred to 0.45 µM nitrocellulose membranes (Hoefer Biotech, Inc., San Francisco, CA) using the electroblotting technique described by Kyhse-Andersen (36). Membranes were probed with the following primary antibodies: mouse monoclonal anti-phosphotyrosine (clone 4G10, Upstate Biotechnology, Inc., Lake Placid, NY), mouse monoclonal anti-pan ERK, which recognizes the 42- and 44-kDa isoforms (Transduction Laboratories, Lexington, KY), polyclonal rabbit anti-phosphorylated-ERK, which recognizes the tyrosine-phosphorylated 42- and 44-kDa ERK isoforms (New England Biolabs Inc., Beverly, MA), polyclonal rabbit anti-phosphorylated-SEK, which recognizes threonine-phosphorylated SEK1 (New England Biolabs), polyclonal rabbit anti-phosphorylated-MEK, which recognizes serine-phosphorylated MEK1/2 (New England Biolabs), and polyclonal rabbit anti-JNK, which is specific for the FLG epitope of JNK and thus recognizes the 46- and 54-kDa isoforms (Santa Cruz Biotechnology, Santa Cruz, CA). Non-immune rabbit IgG or mouse isotype-specific IgG served as a control for the primary antibodies. Appropriate biotinylated secondary antibodies were obtained from Zymed Laboratories, Inc. (San Francisco, CA). After incubation with streptavidin-horseradish peroxidase, blots were developed using enhanced chemiluminescence (Amersham Life Science, Inc.).
ERK was immunoprecipitated from mesangial lysates (500 µg of protein)
with 5 µg of the anti-pan ERK antibody, followed by protein
G-Sepharose. The immunoprecipitated material was run on a
polyacrylamide gel, blotted to nitrocellulose, and probed with the
anti-phosphotyrosine antibody as described above. To determine whether
IL-1 increased JNK tyrosine phosphorylation, 500 µg of protein
from mesangial lysates were immunoprecipitated with 5 µg of
anti-phosphotyrosine, followed by protein A-agarose. After electrophoresis under nonreducing conditions (to avoid interference with the detection of p54 JNK from IgG heavy chains), the
immunoprecipitated material was blotted and probed with the anti-JNK
antibody, as described above.
ERK activity in mesangial cell lysates was
measured as the ability of these lysates to phosphorylate MBP, a
substrate for ERK (37). Briefly, 10 µl of mesangial protein were
incubated with 10 µl of a 2 mg/ml MBP solution in buffer containing
20 mM MOPS, pH 7.2, 25 mM -glycerol
phosphate, 5 mM EGTA, 1 mM
Na3VO4, 1 mM DTT (Upstate
Biotechnology Inc.). A solution containing inhibitors to protein kinase
C, protein kinase A, and calmodulin-dependent kinase was
added to reduce interference from other serine/threonine kinases
(Upstate Biotechnology Inc.). The reaction was started by the addition
of 10 µCi of [
-32P]ATP (6000 Ci/mmol, DuPont NEN) in
buffer containing 75 mM MgCl2 and 500 µM ATP. After 10 min at 30 °C, an aliquot of the
reaction mixture was removed and blotted onto phosphocellulose paper.
The papers were washed several times in 0.75% phosphoric acid before counting in a scintillation counter. After accounting for nonspecific binding and binding of endogenous proteins that were phosphorylated, the results were expressed as pmol of phosphate incorporated into MBP/min/µg of mesangial cell protein.
Mesangial cell
proteins were harvested and JNK activity was assayed (as described by
Cano et al. (38)) after immunoprecipitation of JNK isoforms.
Briefly, cells were lysed in buffer containing 20 mM Hepes,
pH 7.9, 0.3 M NaCl, 2 mM EGTA, 20 mM -glycerophosphate, 1 mM DTT, 0.1% Triton
X-100, 1 mM Na3VO4, 400 µM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin,
and 18 µg/ml aprotinin. After centrifugation and normalization of the
protein content, extracts were mixed with 1.0 µg of polyclonal
anti-JNK antibody (Santa Cruz Biotechnology) overnight at 4 °C. The
JNK·anti-JNK immune complexes were precipitated with protein
G-Sepharose for 1 h, washed several times in kinase buffer (20 mM Hepes, pH 7.9, 2 mM EGTA, 10 mM
MgCl2, 1 mM DTT, 100 µM
Na3VO4, and 25 mM
-glycerophosphate), and resuspended in 40 µl of kinase buffer
containing 2 µg of the fusion protein GST-c-Jun(1-79)
(Santa Cruz Biotechnology), 100 µM ATP, and 3-5 µCi of
[
-32P]ATP. The reaction was allowed to proceed for 20 min at room temperature and was terminated by the addition of
SDS-loading buffer. Phosphorylated GST-c-Jun1-79 was
resolved on a 12% SDS-polyacrylamide gel and then
autoradiographed.
Data are presented as mean ± S. E. Multiple comparisons were examined for significant differences using ANOVA. Individual comparisons were subsequently made using the Bonferroni post-test. A p < 0.05 was considered significant.
The addition of IL-1
to mesangial monolayers resulted in a time-dependent
increase in tyrosine phosphorylation of several proteins. Two proteins
with apparent molecular masses of 41.2 and 52.6 kDa (p41.2 and p52.6,
respectively) were the dominant proteins phosphorylated after
incubation with IL-1
(Fig. 1A). Proteins
of the same molecular mass were tyrosine-phosphorylated when mesangial
cells were treated with diamide, which directly oxidizes intracellular
thiols, or with the membrane-permeable reactive oxygen intermediate,
hydrogen peroxide (Fig. 1B). Because increasing
intracellular oxidation mimicked the effects of IL-1
on protein
tyrosine phosphorylation, we examined whether antioxidants could
inhibit IL-1
-induced tyrosine phosphorylation. As shown in Fig.
2, when mesangial cells were treated with the
antioxidants NAC or DTT before incubation with IL-1
, the tyrosine
phosphorylation of p41.2 and p52.6 was attenuated. Similarly,
diamide-induced tyrosine phosphorylation of p41.2 and p52.6 was blocked
in NAC-treated cells (Fig. 2). These data demonstrate that oxidation
stimulates protein tyrosine phosphorylation in human mesangial cells
and suggest that IL-1
-mediated protein tyrosine phosphorylation may occur through an oxidation-dependent mechanism.
The p41.2 and p52.6 tyrosine phosphoproteins have molecular masses
similar to those reported for the MAP kinases ERK2 and p54 JNK,
respectively (24, 25). Furthermore, the predominant isoforms of ERK and
JNK expressed in human mesangial cells were found to have molecular
masses of 42.1 and 52.8 kDa, respectively (Fig. 3).
These data suggested the possibility that IL-1 treatment of
mesangial cells results in oxidation-dependent
phosphorylation and activation of ERK2 and p54 JNK. This was tested in
the following series of experiments.
ERK2 Is Tyrosine-phosphorylated and Activated in Mesangial Cells by IL-1
To determine whether IL-1 induced
the tyrosine phosphorylation of ERK2, whole cell lysates were probed
with an antibody specific for tyrosine-phosphorylated ERK. As shown in
Fig. 4A, this antibody detected
tyrosine-phosphorylated ERK in lysates from IL-1
-treated mesangial
cells but not from control cells. In most experiments, a single
41.6-kDa isoform of tyrosine-phosphorylated ERK was detected. Occasionally a minor band with an apparent molecular mass of 44.2 kDa
(probably ERK1) was also seen. These results were confirmed by
immunoprecipitating ERK from IL-1
-treated mesangial lysates and
staining the immunoprecipitates with an anti-phosphotyrosine antibody.
This demonstrated a tyrosine-phosphorylated ERK isoform with an
apparent molecular mass of 41.1 kDa (data not shown).
The role of redox status in ERK2 induction by IL-1 was evaluated by
incubating cells with NAC or DTT before IL-1
. These antioxidants
blocked IL-1
-induced tyrosine phosphorylation of ERK2 (Fig.
4A). Furthermore, the oxidizing agents diamide and hydrogen
peroxide reproduced the effects of IL-1
and caused tyrosine phosphorylation of ERK2 (Fig. 4B). This pattern of ERK
phosphorylation in response to IL-1
, oxidants, or antioxidants was
identical to the pattern of induction of the p41.2 tyrosine
phosphoprotein (Fig. 1).
We considered tyrosine phosphorylation of ERK to be an index of enzyme
activation. To verify that IL-1 treatment of mesangial cells
resulted in a functional increase in ERK activity, cell lysates were
tested in a MBP phosphorylation assay. As shown in Table
I, lysates from IL-1
-treated mesangial cells caused a significantly greater phosphorylation of MBP than lysates from untreated cells. Diamide and hydrogen peroxide increased MBP kinase activity similar to IL-1
(Table I). Pretreatment with 2 mM DTT for 30 min completely blocked IL-1
-induced MBP
kinase activity (1.5 ± 0.3 versus 3.8 ± 0.9 pmol
of phosphate incorporated/min/µg of protein; activity from the
control cells was 1.5 ± 0.3, n = 4). Similarly,
MBP kinase activity from cells treated with 20 mM NAC for
30 min followed by IL-1
(2.9 ± 0.5 pmol of phosphate incorporated/min/µg of protein, n = 6) was not
different from control cell activity (2.2 ± 0.4, n = 5) but was significantly less than the activity
from cells treated with IL-1
alone (4.98 ± 0.4, n = 5).
|
Taken together, these data conclusively demonstrate that IL-1 and
oxidizing agents activate ERK2 in human mesangial cells. ERK activation
in response to IL-1
or oxidants has been reported for other cell
types (23, 37, 39-41), consistent with the data presented here. A link
between IL-1
and oxidant-induced ERK activation has not, however,
been established previously. The observation that IL-1
-induced ERK2
tyrosine phosphorylation and activation are inhibitable by antioxidants
provides evidence that reactive oxygen intermediates act as
intracellular signals for IL-1
.
To determine if IL-1 caused tyrosine
phosphorylation of JNK, tyrosine phosphoproteins were
immunoprecipitated from mesangial lysates and probed with anti-JNK
antibody. As shown in Fig. 5, a single isoform of
tyrosine-phosphorylated JNK (molecular mass 56 kDa, nonreducing
conditions) was detected in immunoprecipitates from IL-1
-treated
cells. Phosphorylated JNK was not found in control lysates.
Subsequent experiments were undertaken to assess the role of oxidation
in IL-1-induced JNK activation. JNK was immunoprecipitated from
mesangial lysates, and these immunoprecipitates were used to
phosphorylate the fusion protein GST-c-Jun, a JNK substrate. Minimal
c-Jun phosphorylation was detected using control cell immunoprecipitates, but immunoprecipitates from IL-1
-treated cells
caused significant c-Jun phosphorylation (Fig.
6A). Diamide and hydrogen peroxide also
induced c-Jun kinase activity (Fig. 6B). Immunoprecipitates
from cells exposed to DTT or NAC before IL-1
showed significantly
decreased c-Jun kinase activity compared with cells incubated with
IL-1
alone (Fig. 6A).
These data clearly demonstrate that IL-1 and oxidizing agents
activate JNK in human mesangial cells. IL-1
and oxidant activation of JNK has also been described in other cells (24, 42-44). The observation that IL-1
-induced JNK activity can be attenuated by
antioxidants provides evidence that reactive oxygen species may mediate
the effects of IL-1
on JNK activation in human mesangial cells. This
conclusion is supported by a recent study that showed antioxidant
inhibition of IL-1
-induced JNK activation in bovine chondrocytes
(43).
The activation of ERK is dependent on an
upstream Tyr/Thr kinase designated as MEK (45). In order for MEK to
phosphorylate and activate ERK, MEK must itself be
serine-phosphorylated (46, 47). This portion of the ERK activation
cascade was analyzed by immunoblotting mesangial lysates with an
antibody specific for serine-phosphorylated (and presumably activated)
MEK. As shown in Fig. 7, IL-1, hydrogen peroxide, and
diamide each induced MEK phosphorylation. The phosphorylation of MEK
correlated with its activation because PD 098059, a specific inhibitor
of MEK activation (35), prevented IL-1
- and oxidant-induced ERK
tyrosine phosphorylation (data not shown). This is consistent with a
recent study that showed oxidant induction of ERK was mediated in part through MEK activation (40). These data support the conclusion that
IL-1
activates ERK through an oxidation-dependent
pathway. Furthermore, the results show that an IL-1
-induced redox
signal acts upstream of ERK tyrosine phosphorylation, at least at the level of MEK activation, if not earlier in the cascade. These data also
suggest that redox signals may regulate phosphorylation events other
than protein tyrosine phosphorylation.
Like ERK, JNK phosphorylation and activation are dependent on an
upstream Tyr/Thr kinase. One potential upstream JNK kinase is SEK1,
which itself must be serine/threonine-phosphorylated for activity (48,
49). Immunoblotting mesangial lysates for the threonine-phosphorylated
form of SEK1 demonstrated phosphorylated SEK1 in lysates of
IL-1-treated cells but not in control cells (Fig. 7). In contrast,
hydrogen peroxide and diamide did not induce phosphorylation of SEK1
(Fig. 7). This suggests that IL-1
and exogenous oxidants activate
JNK by different mechanisms. A similar discrepancy was seen in
neutrophils in which exogenous oxidants activated MAP kinase, but
endogenously derived oxidants did not mediate MAP kinase activation by
N-formyl-methionyl-leucyl phenylalanine (40). Although
exogenous oxidants may conceivably activate JNK directly, a recent
report suggests that hydrogen peroxide could not directly activate JNK
in vitro (43). Alternatively, exogenous oxidants may act
through a different upstream regulatory element than SEK1 (50). The
observation that exogenous oxidants do not mimic the ability of IL-1
to phosphorylate (and presumably activate) SEK1 challenges the idea of
a redox signal mediating IL-1
-induced JNK activation. This, however,
is difficult to reconcile with the antioxidant inhibition of
IL-1
-induced JNK activation (this study and Ref. 43). A potential
explanation of these results is that in contrast to exogenous oxidants,
endogenously derived reactive oxygen species generated during IL-1
treatment do contribute to the activation of SEK1. This scenario is
supported by the observation that IL-1
-induced SEK1 phosphorylation
was prevented in mesangial cells pretreated with DTT (data not
shown).
In summary, this report demonstrates that IL-1 and exogenous
oxidizing agents activate ERK2 in human glomerular mesangial cells. For
both IL-1
and oxidants, activation of ERK2 proceeds through MEK.
These data, along with the finding that IL-1
-induced ERK2 activation
is attenuated by antioxidants, provide strong evidence that IL-1
activates ERK through an oxidation-dependent pathway.
IL-1
and exogenous oxidizing agents also activate p54 JNK but appear
to signal through different upstream pathways. Exogenous oxidants act
independently of SEK1, whereas IL-1
treatment results in SEK1
phosphorylation and presumably activation. Thus, IL-1
may activate
JNK through an oxygen radical independent pathway. Alternatively, since
antioxidants block JNK activation and SEK1 phosphorylation in response
to IL-1
, endogenously derived oxygen radicals may serve as signal
transduction elements for IL-1
in the JNK cascade. The implications
of these results for human renal disease remain to be elucidated, but
the data clearly suggest that at the molecular level MAP kinases may
mediate the effects of IL-1
on glomerular cells. Additionally, this
report illustrates a potential mechanism through which exogenous oxygen
radicals, released during glomerular inflammation, can influence renal
cellular function. Finally, the fact that endogenous reactive oxygen
species may act as second messengers in the MAP kinase activation
pathway identifies a potential target for therapeutic intervention in glomerular disease.