| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
-Induced Apoptosis in Rat Fetal Brown Adipocytes1
Departamento de Bioquímica y Biología Molecular II, Instituto de Bioquímica, Centro Mixto del Consejo Superior de Investigaciones Científicas y de la Universidad Complutense de Madrid (A.V., J.J.V., C.R., M.B., A.P.); Centro de Citometría de Flujo y Microscopía Confocal, Universidad Complutense de Madrid (A.M.A.); Facultad de Farmacia, Universidad Complutense, Ciudad Universitaria, 28040 Madrid, Spain
Address all correspondence and requests for reprints to: Dr. Almudena Porras, Departamento de Bioquímica y Biología Molecular II, Instituto de Bioquímica, Centro Mixto del Consejo Superior de Investigaciones Científicas y de la Universidad Complutense de Madrid, Facultad de Farmacia, 28040 Madrid, Spain.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TNF) induces apoptosis and cell growth
inhibition in primary rat fetal brown adipocytes. Here, we examine the
role played by some members of the mitogen-activated protein kinase
(MAPK) superfamily. TNF activates extracellular regulated kinase-1/2
(ERK1/2) and p38MAPK. Inhibition of p38MAPK by either SB203580 or
SB202190 highly reduces apoptosis induced by TNF, whereas ERK
inhibition potentiates it. Moreover, cotransfection of an active MKK3
mutant and p38MAPK induces apoptosis. p38MAPK inhibition also prevents
TNF-induced cell cycle arrest, whereas MEK1 inhibition enhances this
effect, which correlates with changes in proliferating cell nuclear
antigen expression, but not in cyclin D1.
c-Jun and activating transcription factor-1 are potential downstream
effectors of p38MAPK and ERKs upon TNF treatment. Thus,
TNF-induced c-Jun messenger RNA expression requires ERKs activation,
whereas p38MAPK inhibition enhances its expression. In addition,
TNF-induced activating transcription factor-1 phosphorylation is
extensively decreased by SB203580. However, TNF- induced NF-
B
DNA-binding activity is independent of p38MAPK and ERK activation. On
the other hand, C/EBP homology protein does not appear to
mediate the actions of TNF, because its expression is almost
undetectable and even reduced by TNF.
Finally, although TNF induces c-Jun N-terminal kinase (JNK)
activation, transfection of a dominant negative of either JNK1 or JNK2
had no effect on TNF-induced apoptosis. These results suggest that
p38MAPK mediates TNF-induced apoptosis and cell cycle arrest,
whereas ERKs do the opposite, and JNKs play no role in this process of
apoptosis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tumor necrosis factor- (TNF) may play an important role as a
negative regulator of this tissue by different mechanisms. TNF
inhibits differentiation (7) in these cells as it does in
white adipocytes and in some cell lines, such as 3T3-L1 cells, that can
be induced to differentiate into adipocytes (8, 9). TNF
can also induce reversal of adipocytic differentiation
(8). Additionally, we have demonstrated that TNF
inhibits cell growth and induces apoptosis in primary rat fetal brown
adipocytes (10). In mature brown adipocytes, TNF (in
the presence of cycloheximide) also induces a process of apoptosis and
the number of apoptotic cells is greater in obese animals than in lean
rats (11). Therefore, by this mechanism, TNF might play
a relevant role in the control of the number of brown adipocytes during
the perinatal period and/or in obese animals. A similar regulation
might take place in white adipocytes, where TNF is also able to
induce apoptosis (12). However, the signaling pathways
mediating these TNF actions are not known.
TNF activates different signaling pathways in different cell types,
such as extracellular regulated kinases (ERKs), c-Jun N-terminal
kinases (JNKs), p38 mitogen-activated protein kinases (p38MAPKs),
nuclear factor-B (NF-B), and caspases (13, 14, 15, 16). ERKs
are activated by many signals (15, 16), particularly by
mitogens, and their activation is necessary for proliferation
(17, 18, 19) or differentiation (18, 20),
depending on the cell type, and could also play a role in survival
(21). p38MAPK and JNKs are mainly activated by
proinflammatory cytokines such as TNF and cellular stresses
(22, 23, 24) in different cell types, although recently
p38MAPK has been demonstrated to play a role in a number of cellular
functions, such as differentiation, cell motility, developmental
processes, and survival. p38MAPK was identified as a kinase induced by
stress signals (25), and it is now referred to as
p38MAPK or p38MAPK. Other isoforms have been described that belong
to the p38 subfamily of MAPKs, constituted by different members:
p38, p38ß, p38
(ERK6/stress-activated protein kinase 3),
and p38
(reviewed in Ref. 24).Among these different
members, p38MAPK has been more widely characterized as have their
different cellular effects. p38MAPKs have been implicated both as
positive (21, 26, 27, 28) and negative (29, 30)
regulators of apoptosis. They have also been reported to be involved in
cell growth regulation (31, 32). On the other hand, the
involvement of JNKs in TNF-induced apoptosis or even in apoptosis
induced by other stimuli has been controversial (21, 26, 33, 34, 35).
In this work we have analyzed the roles played by p38MAPK and ERKs in
TNF-induced apoptosis and cell cycle arrest in rat fetal brown
adipocytes. We present here evidence that p38MAPK is activated by
TNF and is involved in TNF-induced apoptosis and cell growth
inhibition. In contrast, ERKs, which are also activated by TNF,
mediate cell growth and survival and, as a consequence, attenuate the
effects of TNF. As inhibition of p38MAPK did not block
TNF-induced apoptosis in all the cells, the participation of JNK in
this process was also explored, but no role was found.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(1–10 ng/ml), IGF-I (2.5
nM), insulin (20 nM), or combinations thereof
for different time periods as indicated. To inhibit ERKs, cells were
pretreated with the MEK1 inhibitor PD98059 (no. 513000,
Calbiochem, La Jolla, CA) at concentrations between 5–50
µM for 1 h before the addition of signals.
Similarly, to inhibit p38MAPK, cells were pretreated either with the
specific inhibitor SB203580 (1–5 µM; SmithKline Beecham,
Harlow, UK) or with SB202190 (2 µM; no. 559388,
Calbiochem).
Culture of MB4.8.2 brown adipocyte derived cell line and
transfection assays
The MB 4.8.2 brown adipocyte-derived cell line was grown in DMEM
supplemented with 10% FBS. To determine apoptosis, cells were
maintained in DMEM with 3% FBS and in the presence or absence of
TNF (10 ng/ml) for 48 h. To measure JNK activity, cells were
serum starved overnight and then triggered with TNF for 10 min.
Transient transfections were carried out using the Gene PORTER transfection reagent from Gene Therapy Systems (catalog no. T201007, San Diego, CA) following the protocol supplied by the manufacturer. Cells were transfected with an active MKK3 mutant (MKK3DD) plus a p38MAPK wild-type construct and with different JNKs constructs: JNK1 wild-type, JNK2 wild-type, and a dominant negative of JNK1 (JNK1APF) and JNK2 (JNK1APF) (36), subcloned in a green fluorescent protein (GFP) expression vector from CLONTECH Laboratories, Inc. (Palo Alto, CA).
Transfection efficiencies ranged between 50–70%, as determined by the ß-galactosidase histochemical staining assay or by quantification of GFP expression by cytometry and fluorescent microscopy.
Flow cytometric analysis
Analysis of DNA content, cell cycle, and proliferating cell
nuclear antigen (PCNA) content of the cells in the different phases of
the cell cycle was performed in a FACScan flow cytometer (Becton Dickinson and Co., San Jose, CA). DNA was stained with propidium
iodide (PI) using the Bio-Rad Laboratories, Inc. reagent
kit (Kinesis 50, 470-0023, Richmond, CA), following the manufacturer’s
protocol. When PCNA content was analyzed in parallel to the cellular
DNA content, the Bio-Rad Laboratories, Inc., reagent kit
containing PI and an anti-PCNA antibody (PCNA/Kinesis 50, 470-0043) was
used. Measurements were carried out using a Double Discriminator Module
to discriminate doublets. Ten thousand cells were acquired per sample.
Then, the percentage of cells with DNA content lower than 2C was
calculated as well as the percentage of cells in the
G0/G1, S, and
G2/M phases of the cell cycle, using Multicycle
software (Phoenix Software, Mountain View, CA). When PCNA and
cellular DNA contents were simultaneously analyzed, the percentage of
cells expressing PCNA with different DNA contents was also calculated
using CellQuest software (Becton Dickinson and Co.,
Mountain View, CA).
The presence of phosphatidylserine in the outer layer of the plasma membrane (an early feature of apoptosis) was detected by specific binding to annexin V using the annexin V-Fluos kit from Roche Molecular Biochemicals (catalog no. 1828681, Indianapolis, IN). Cells were incubated with PI and/or annexin V and analyzed by flow cytometry. The percentage of cells positive for annexin V and negative or low positive for PI, considered apoptotic, and the percentage of those positive for annexin V and PI, considered necrotic, were determined.
Analysis of DNA fragmentation
DNA from the extranuclear fraction was isolated as previously
described (10). Then DNA was electrophoresed in a 1.5%
agarose gel.
Western blot analysis
Total cell extracts were obtained in a buffer containing 25
mM HEPES (pH 7.5), 0.3 M NaCl, 1.5
mM MgCl2, 0.2 mM EDTA,
0.1% Triton X-100, 0.5 mM dithiothreitol, 20
mM ß-glycerophosphate, 1 mM
Na3VO4, leupeptin (2
µg/ml), and 1 mM phenylmethylsulfonylfluoride.
Active ERK (ERK1/2) levels were quantified in total cell extracts by
Western blot analysis using an antiphospho-ERKs antibody from
Promega Corp. (V6671, Madison, WI). After stripping of the
membranes, total ERK levels were quantified by subsequent incubation
with the anti-ERK1/2 antibody from Upstate Biotechnology, Inc. (06–182, Lake Placid, NY) as previously described
(17, 37). Similarly, active p38MAPK was quantified in
total cell extracts using an antiphopho-p38MAPK antibody from New England Biolabs, Inc. (no. 9211S, Beverley, MA), and total
p38MAPK levels were quantified using an anti-p38MAPK antibody from
Santa Cruz Biotechnologies, Inc. (sc-535, Santa Cruz,
CA).
PCNA, cyclin D1, CHOP (C/EBP homology protein) (GADD153), and phospho-cAMP response element-binding protein (phospho-CREB)/activating transcription factor-1 (ATF1) were also quantified in total cell extracts by Western blot analysis, using the following antibodies: a monoclonal anti-PCNA antibody from Roche (1486772), an antihuman cyclin D1 from PharMingen (66271A, San Diego, CA), a polyclonal anti-GADD153 from Santa Cruz Biotechnologies, Inc. (sc-793), and a phospho-CREB (Ser133) antibody from New England Biolabs, Inc. (9191S).
Blots were developed using the ECL system (Amersham Pharmacia Biotech, Arlington Heights, IL). Autoradiograms were quantified using computer-assisted densitometry (Molecular Dynamics, Inc., Sunny- vale, CA).
MAPKAPK2 kinase assay
To measure MAPKAPK2 kinase activity, cell lysates were obtained
as described for phospho-p38MAPK Western blot analysis. Then, they were
immunoprecipitated with an anti-MAPKAPK2 antibody from StressGen
Biotechnologies Corp. (Victoria, British Columbia, Canada)
(KAP-MA015E). Immunoprecipitates were washed twice with the lysis
buffer and twice with a kinase buffer containing 25 mM
HEPES (pH 7.4), 25 mM MgCl2, 25
mM ß-glycerophosphate, 100 µM sodium
orthovanadate, and 2 mM dithiothreitol. Then, samples were
incubated at 30 C for 30 min with the kinase buffer containing 100
µM cold ATP, 10 µCi
[-32P]ATP, and 2 µg/assay of human
recombinant 27-kDa heat shock protein (Hsp27) protein (StressGen
Biotechnologies Corp., SPP-715) as substrate. The reaction was stopped
by adding Laemmli sample buffer. Phosphorylated Hsp27 was visualized by
autoradiography after electrophoresis on 12% SDS-polyacrylamide
gels.
[3H]Thymidine incorporation assay
Cells were triggered for 4 h with IGF-I, insulin, TNF,
or combinations thereof in the presence or absence of the MEK or
p38MAPK inhibitors as indicated and then incubated for 24 h with a
mix of [3H]thymidine and cold thymidine at
final concentrations of 1 µCi and 20 µM, respectively.
Radioactivity present in trichloroacetic acid-insoluble material was
quantitated in a ß-counter.
JNK assay
JNK activity was determined in cell lysates, obtained as
described for Western blot analysis. A pull-down of JNKs was performed
with glutathione-S-transferase (GST)-c-Jun 79 protein
(containing the first 79 amino acids fused to GST) provided by Dr.
Silvio Gutkind. Then, protein complexes were washed 3 times with PBS
containing 1% Nonidet P-40 and 2 mM sodium
vanadate, once with 100 mM Tris (pH 7.5) and 0.5
M LiCl, and once with the kinase buffer [12.5
mM 4-morpholine-propanesulfonic acid (pH
7.5), 12.5 mM ß-glycerophosphate, 7.5
mM MgCl2, 0.5
mM EGTA, 0.5 mM NaF, and
0.5 mM sodium vanadate]. The kinase assay was
carried out in this kinase buffer containing 0.3 µCi
[-32P]ATP and 20 µM
cold ATP. The reaction was stopped by adding Laemmli sample buffer.
Phosphorylated c-Jun was visualized by autoradiography after
electrophoresis on 10% SDS-polyacrylamide gels.
RNA extraction and Northern blot analysis
Northern blot analysis and total RNA extraction were performed
as previously described (17). Blots were hybridized with a
c-Jun probe given by Dr. M. Yaniv, then stripped and rehybridized with
an 18S ribosomal probe to normalize. Radioactivity was quantitated in a
Fuji Photo Film Co., Ltd. BAS-1000 apparatus (Tokyo,
Japan).
NF-B gel mobility shift assay
Nuclear extracts were obtained essentially as described
previously (38). Then, these extracts were used for the
NF-B mobility shift assay, which was performed as previously
described (38), using the double stranded oligonucleotide
corresponding to the B distal motif present in the murine iNOS
promoter: TGCTAGGGGGATTTTCCCTCTCT CTGT.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
activates ERKs (0.5–20 ng/ml) for 10 min. The levels of active ERKs,
analyzed by phospho-MAPK Western blot, increased after TNF
treatment, reaching a maximum with 10 ng/ml (Fig. 1A
).
|
(10 ng/ml) and
reached a peak at 10 min (Fig. 1B
). Then, ERKs activity decreased after
30 min.
This TNF-induced increase in ERK activity was completely blocked by
pretreatment with the specific MEK1 inhibitor, PD98059, using a dose of
5–50 µM (Fig. 1C). Hence, we chose a dose of 20
µM PD98059 to study the role of ERKs in the TNF
actions, which inhibited ERKs and had no effect on the activity of
other kinases, such as JNKs or p38MAPK (data not shown) in our cell
system.
TNF activates p38MAPK
p38MAPK can be activated by proinflammatory cytokines such as
TNF (23, 24, 25) in different cell types. Hence, we
analyzed the levels of active p38MAPK in brown adipocytes upon
treatment with TNF by Western blot, using an anti-phospho-p38MAPK
antibody.
As shown in Fig. 2A, after 5 min of
treatment with a dose of TNF as low as 1 ng/ml, p38MAPK was
activated, reaching a maximum with 5–20 ng/ml. Thus, 10 ng/ml TNF
would be an optimal dose for our experiments, taking into account that
maximum activation of p38MAPK and ERKs was obtained with this dose.
|
was also performed.
p38MAPK was activated very quickly, reaching a maximum at 5 min that
was maintained until 10 min and decreasing thereafter (Fig. 2B).
To study the role played by p38MAPK in TNF treatment of brown
adipocytes, this pathway could be inhibited with the specific p38MAPK
inhibitor, SB203580. This inhibitor is specific for the isoform at
low doses, but at higher doses it also inhibits p38ß. It acts by
competition with ATP (39); therefore, to detect p38MAPK
inhibition, it is necessary to look for inhibition of downstream
targets such as MAPKAPK2 (25). Treatment of the cells
with TNF for 5 min activated MAPKAPK2, and this activation was
completely blocked when cells were previously pretreated with doses of
2–5 µM SB203580 (Fig. 2C) or 2 µM SB202190
(data not shown). Thus, to analyze the effect of p38MAPK cascade
inhibition on TNF-induced apoptosis and cell growth arrest, we used
5 µM SB203580 (and 2 µM SB202190 to confirm
apoptosis data). With this dose of inhibitor, the TNF- induced
p38MAPK cascade was specifically blocked; other MAPK pathways, such as
ERKs and JNKs, were not affected (data not shown).
Roles of p38MAPK and ERKs in cell growth inhibition induced by
TNF
As ERKs and p38MAPK can regulate the cell cycle, we
analyzed their roles in TNF-induced cell cycle
arrest in brown adipocytes using the p38MAPK inhibitor, SB203580, and
the MEK1, PD98059. To do this, flow cytometric analysis of the cell
cycle was performed (Fig. 3A). Treatment
of brown adipocytes with TNF for 24 h induced a significant
decrease in the percentage of cells in the proliferative phases of the
cell cycle compared with that in control cells (P <
0.05). However, pretreatment of these cells with the p38MAPK inhibitor
abolished this effect. Moreover, the significant decrease observed in
the percentage of cells in S+G2/M phases of the
cell cycle in cells treated with IGF-I/insulin plus TNF compared
with that in IGF-I/insulin-treated cells (P < 0.01 and
P < 0.05, respectively) was reversed when p38MAPK was
inhibited. In contrast, MEK inhibition potentiated the effect of
TNF, significantly decreasing the percentage of cells in
S+G2/M phases of the cell cycle under basal
conditions (P < 0.05) and in cells treated either with
IGF-I or insulin plus TNF (P < 0.05 and
P < 0.001, respectively). In addition, the
IGF-I/insulin effect to increase the percentage of cells in the
proliferative phases of the cell cycle was totally abolished by
inhibition of ERKs.
|
. TNF decreased significantly
[3H]thymidine incorporation, and this effect
was partially abolished by pretreatment with the p38MAPK inhibitor
(P < 0.05 compared with TNF-treated cells), whereas
MEK1 inhibition additionally decreased DNA synthesis (P
< 0.01). TNF also significantly decreased IGF-I- and
insulin-induced DNA synthesis by 80% and 50–60%, respectively. This
TNF effect was not observed when cells were pretreated with the
p38MAPK inhibitor, and it was significantly potentiated by pretreatment
with PD98059 (P < 0.01). ERKs cascade inhibition also
blocked the effects of IGF-I and insulin. Therefore, all of these data
suggest that p38MAPK mediates TNF-induced cell cycle arrest, whereas
ERKs mediate cell growth in these cells.
In fibroblasts, regulation of cell growth by p38MAPK and ERKs pathways
correlates with changes in cyclin D1 expression (31).
However, no significant changes in the level of cyclin D1 (Fig. 3C)
were observed in brown adipocytes upon TNF treatment (at different
time periods) in either the absence or presence of the p38MAPK and MEK
inhibitors. Different from this, the level of another cell
cycle-regulated protein, PCNA, decreased after treatment with TNF
(Fig. 3D). This effect was reversed by pretreatment of the cells
with the p38MAPK inhibitor, and it was potentiated by MEK inhibition.
Therefore, PCNA expression is regulated by p38MAPK and ERKs pathways,
and it may be involved in the regulation of cell growth under these
conditions.
TNF-induced apoptosis is partially blocked by inhibition of
p38MAPK, but it is potentiated by MEK inhibition
To study whether p38MAPK and ERKs played any role in
TNF-induced apoptosis in brown adipocytes, cells were treated with
TNF in the presence or absence of specific inhibitors of these
pathways, and apoptosis was quantified by flow cytometry after staining
with propidium iodide (Table 1).
Treatment with TNF for 48 h induced a 2.7-fold increase in the
percentage of cells with DNA content lower than 2C, which was
reduced by about 36% (compared with that in TNF-treated cells) when
p38MAPK was inhibited by SB203580. As SB203580 alone slightly increased
apoptosis in control cells, protection from the TNF-induced
apoptosis with this inhibitor would be much higher. In addition,
blockade of p38MAPK with another inhibitor, SB202190, led to a similar
effect (Table 1). However, pretreatment with the MEK inhibitor enhanced
the TNF effect, leading to a 3.5-fold increase. Therefore, these
data suggest that p38MAPK is involved in the mediation of apoptosis,
whereas ERK activation might have a protective effect.
|
(Fig. 4A), whereas MEK inhibition
had the opposite effect. Similarly, the percentage of cells considered
apoptotic (positive for annexin V and negative or low positive for PI)
was decreased in TNF-treated cells pretreated with the p38MAPK
inhibitor and increased in cells pretreated with the MEK inhibitor
(Fig. 4B).
|
in brown adipocytes. To confirm this by a
different approach, transient transfections with wild-type p38MAPK plus
an active mutant of MKK3 (MKK3DD), a kinase upstream of p38MAPK, were
carried out in a brown adipocyte-derived cell line (MB 4.8.2)
previously generated (40), as transfection efficiency is
very low in primary brown adipocytes. The percentage of cells with DNA
content lower than 2C was much higher in cells transfected with
the active mutant of MKK3 (MKK3DD) plus wild-type p38 MAPK (Fig. 5B) than in cells transfected with the
empty vector (Fig. 5A) in both untreated cells and cells treated with
TNF. The increase in the number of apoptotic cells in transfected
cells correlated with an increase in the level of active p38MAPK (Fig. 5, upper panel). These data demonstrated that p38MAPK
activation can induce apoptosis in brown adipocytes.
|
induces apoptosis in nonproliferating brown adipocytes:
effect of p38MAPK and ERK inhibition induces apoptosis in brown
adipocytes, which do not express PCNA and, therefore, are not
proliferating (10). Thus, to establish a possible
relationship between the roles played by p38MAPK and ERKs in cell cycle
and apoptosis, we simultaneously measured DNA content and PCNA
expression by flow cytometry (Table 2).
Under control conditions, half of the cells dying expressed PCNA.
TNF treatment highly increased the percentage of apoptotic cells;
most of them were negative for PCNA expression, suggesting that TNF
induces cell death in nonproliferating cells, as previously
demonstrated (10). Pretreatment with SB203580 did not
change this effect significantly, whereas inhibition of MEK with
PD98059 induced a decrease in the percentage of cells negative for PCNA
expression, and most of the cells expressed PCNA. Therefore, although
inhibition of p38MAPK decreases apoptosis induced by TNF, cells that
are not proliferating still die. In contrast, ERKs inhibition increases
the number of apoptotic cells compared with TNF, and most of these
cells progress through the cell cycle.
|
B activation, c-Jun
messenger RNA (mRNA) expression, CHOP protein levels, and CREB/ATF1
phosphorylationB, can be modulated by p38MAPK (41, 42, 43, 44, 45).
On the other hand, ERKs have been shown to regulate the CREB/ATF family
(45) as well as other transcription factors, such as the
c-Jun family (46). Hence, they are putative targets of
p38MAPK and/or ERKs.
c-Jun is one of the components of the transcription factor
activating protein-1 and also forms Jun-ATF dimers, regulating gene
expression and different cellular events, such as proliferation and
apoptosis. Its regulation is very complex and includes changes in c-Jun
expression and in protein phosphorylation. In brown adipocytes, we show
in Fig. 6A, that treatment with TNF
for 30 min induced an increase in c-Jun mRNA level, which was prevented
by pretreatment of the cells with PD98059, but was enhanced by
SB203580.
|
B, plays a protective role in some
processes of apoptosis (47, 48, 49), and its activity can be
regulated by p38MAPK (29, 43, 50, 51) and ERKs
(52). In brown adipocytes, we found that TNF increased
NF-B binding activity, but neither pretreatment with SB203580 nor
that with PD98059 modified this effect (Fig. 6B).
The transcription factor, CHOP/GADD153, could be another possible
target for p38MAPK, for it can be regulated by p38MAPK-mediated
phosphorylation (42) and by changes in gene expression
(43). As shown in Fig. 7A, the CHOP protein level was very low in control cells and was
additionally decreased by TNF at 24 and 48 h (data not shown).
p38MAPK inhibition and MEK inhibition slightly reduced this decrease.
Therefore, it appears that CHOP could not be a mediator of
TNF-induced apoptosis, as it was almost undetectable in cells after
TNF treatment.
|
). No
change was observed after treatment of the cells with TNF in either
the presence or absence of these inhibitors. However, TNF induced
the phosphorylation of ATF-1 (Fig. 7B), which was highly reduced by
inhibition of p38MAPK, but not by ERK inhibition. Therefore, ATF1 might
be a mediator of the TNF effects produced via p38MAPK.
TNF activates JNKs: role of JNK1/2 in TNF- induced
apoptosis
TNF activates JNKs in different cell types. In brown
adipocytes, we studied JNK activity after treatment with
different doses of TNF (0.5–20 ng/ml) for 10 min, and the
results are shown in Fig. 8A. TNF
treatment induced an increase in JNK activity, reaching a maximum with
5 ng/ml.
|
(10 ng/ml) and
reached a peak of maximum activation at 10–15 min, decreasing
afterward (Fig. 8B).
As the TNF-induced apoptosis in brown adipocytes was highly reduced
by blockade of p38MAPK, but was not completely abolished, JNKs might
play a role. Therefore, we determined whether JNKs could be involved in
mediation of this process of apoptosis. To do this, JNK1 and JNK2
wild-type constructs and dominant negative APF mutants
(36), subcloned in a GFP expression vector, were
transfected in the brown adipocyte-derived cell line (MB 4.8.2)
(40). The percentage of cells with DNA content lower than
2C did not change when wild-type JNK1 was overexpressed in
either control or TNF-treated cells. However, in cells
overexpressing wild-type JNK2, a small increase in percentage was
produced, which was additionally increased by TNF similar to the one
obtained in cells transfected with the vector (data not shown). This
would suggest that JNK2, but not JNK1, might play a role in this
process of apoptosis. However, the expression of the dominant negative
JNKs mutants was unable to block TNF-induced apoptosis (Fig. 9B), although TNF-induced JNK
endogenous activity was highly reduced by these JNKs mutants (Fig. 9A),
indicating that this pathway is not involved in mediation of
TNF-induced apoptosis in brown adipocytes. All of these data were
confirmed by flow cytometric analysis of the percentage of hypodiploid
cells, only in the cells expressing GFP and therefore expressing the
transfected JNKs constructs (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
induces apoptosis and cell cycle arrest in rat fetal brown
adipocytes in primary culture (10). We have analyzed here
the roles played by p38MAPK and ERKs as possible mediators of these
TNF actions as well as the role played by JNKs in this process of
apoptosis. We have demonstrated that TNF activates ERK1/2 in these
cells with a peak at about 10 min. This activation is mediated by MEK1,
as no ERK activity was detected when cells were pretreated with the
MEK1-specific inhibitor, PD98059, at doses of 5–20 µM,
which are high enough to completely inhibit MEK1, but not MEK2
(53). Similarly, TNF activates ERKs in other cell
types, such as 3T3-L1 cells (54), and in mouse
macrophages, where MEK1 is also upstream of ERKs (15). We
have also shown here that p38MAPK and its downstream kinase, MAPKAPK2,
were activated by TNF, with a peak at 5 min. MAPKAPK2 activation was
totally blocked by the p38MAPK-specific inhibitor, SB203580, at doses
of 2–5 µM, which do not affect other kinases (data not
shown). However, in 3T3-L1 cells, the activation of p38MAPK by TNF
seems to be different depending on the differentiation state. Thus, in
undifferentiated 3T3-L1 cells, TNF highly activated p38MAPK
(55), whereas in 3T3-L1 adipocytes (white adipocytes),
either a small increase in this activity upon TNF treatment
(55) or even no activation at all (52) was
found. Regarding this, the strong activation of p38MAPK induced by
TNF in fetal brown adipocytes is not surprising, because these cells
are not fully differentiated and are still able to proliferate and
differentiate in response to the appropriate signals. In addition,
brown adipocytes can have a different response to TNF than white
adipocytes due to cell type differences. However, what is probably more
likely is that discrepancies can be due to differences in the
differentiation state of the cells. Among the different isoforms of
p38MAPK described to date, it is very likely that the isoform activated
by TNF in brown adipocytes could be p38MAPK, because the
anti-p38MAPK antibody used for the Western blot was designed against a
peptide with a sequence present in p38MAPK and recognizes a band
corresponding to that detected by the antiphospho-p38MAPK antibody.
Moreover, MAPKAPK2, which was described to be preferentially activated
by p38, but not by p38ß, p38, or p38 (25), was
inhibited by a dose of SB203580 that mainly affected the p38
isoform. Higher doses can also inhibit the p38ß isoform. Therefore,
most of the interfering SB203580 effects on TNF actions that we have
described here might be due to inhibition of the p38 isoform,
although participation of other p38 isoforms, such as p38ß, cannot be
totally ruled out.
We found that ERKs and p38MAPK play opposite roles in TNF-induced
apoptosis in brown adipocytes. p38MAPK mediates this process, whereas
ERKs seem to play a survival role. Regarding this, the DNA ladder shown
in Fig. 4A appears to indicate that inhibition of MEK doubles the
TNF-induced DNA fragmentation, whereas data from Table 1, in which
hypodiploid peaks are shown, indicate a 30% increase. However, it
should be pointed out that the data presented in Table 1 and Fig. 4B
can be considered quantitative, whereas the DNA ladder presented in
Fig. 4A is more qualitative than quantitative. Thus, these data do not
reflect the precise percentage of cells undergoing apoptosis in each
case, although an increase or decrease in DNA laddering always
correlated with an increase or decrease in the percentage of cells with
DNA content lower than 2C (Table 1). On the other hand, both
quantifications have different meanings. DNA laddering in the gel
reflects the extranuclear fragmented DNA accumulated during the
treatment, whereas cytometric analysis of DNA content allows us to
quantify the number of apoptotic cells at a particular moment. However,
DNA laddering demonstrates that the process of cell death taking place
is apoptosis, instead of necrosis. On the other hand, although the
effects of the p38MAPK and ERKs inhibitors were not dramatic, they were
highly significant for asynchronized cells maintained in primary
culture, particularly if the high number of experiments carried out was
considered. In addition, the role played by p38MAPK in this process of
apoptosis was confirmed by the marked increase observed in basal and
TNF-induced apoptosis in cells transiently transfected with an
active mutant of MKK3 (MKK3DD) plus wild-type p38MAPK (Fig. 5). This
is in agreement with the increased apoptosis observed in 3T3-L1 cells
transfected with a constitutively active MKK6 (56). In
addition, the roles proposed here for p38MAPK and ERKs agree with those
described for PC12, in the apoptosis triggered by nerve growth factor
withdrawal (21). In this case, it seems that the mediator
of this process of apoptosis is Fas ligand, the expression of which is
induced via p38MAPK and JNK (57). Regarding TNF, early
activation of p38MAPK and JNK was proposed to be necessary for the
survival of L929-cyt16 cells, whereas ERKs activation might not play
any role (58). Therefore, this role suggested for p38MAPK
is opposite that we propose here. However, it should be pointed out
that there are many differences between the process of apoptosis
induced by TNF in L929-cyt16 cells and the one induced in brown
adipocytes, which can explain our different results. Thus, in L929,
apoptosis induced by TNF, requires the presence of actinomycin D,
and it is produced at about 6 h. In brown adipocytes, it occurs at
48 h and does not require actinomycin D. In addition, a different
kinetic for p38MAPK was found: a second peak of p38MAPK (and JNK)
activation at 6 h was produced in L929-cyt 16, coincident with
apoptosis, that was not present in brown adipocytes. The researchers
argue that the first peak of activation of p38MAPK and JNK is necessary
for protection against apoptosis by TNF, but it is probably not
enough, so the second peak could be important as well. Thus, the first
peak of activation occurs regardless of whether the cells undergo
apoptosis, whereas the second one only occurs in cells undergoing
apoptosis. On the other hand, the concentration of SB203580 used in the
experiments was higher than that we used here. We used 5
µM SB203580, whereas the potentiation of TNF-induced
apoptosis in L929-cyt16 was only clearly detected with a dose of 30
µM, which can potentially inhibit p38 and p38ß. As a
consequence, the participation of p38ß cannot be excluded. In
addition, L929-cyt16 cells are cells completely different from brown
adipocytes: they are a different cell type, they are not primary cells,
and they stably express a chimeric receptor consisting of the
extracellular transmembrane domains of murine CD4 fused to the
cytoplasmic domain of murine Fas. Hence, it is hard to compare the
results obtained in these cells with our results. Thus, the enhancement
of TNF-induced apoptosis can be due to cell type differences that
lead to different kinetics of p38MAPK and JNK activation and/or to the
simultaneous inhibition of p38 and p38ß by SB203580. According to
this, other studies suggest different roles for the p38 and p38ß
isoforms; p38 appears to induce apoptosis while p38ß enhances
survival (59, 60). In addition, in other processes of
apoptosis, p38s have been implicated as both positive (21, 26, 27, 28) and negative (29, 30) regulators of
apoptosis. Hence, all of these discrepancies may be due to cell type
and/or stimuli differences or even to participation of different p38
isoforms.
Because the process of apoptosis induced by TNF in brown adipocytes
was not totally prevented by inhibition of p38MAPK, participation of
other signaling pathways, different from p38MAPK, cannot be excluded.
In fact, in some processes of apoptosis, it has been proposed that both
p38MAPK and JNK are mediators of the process (21, 56). In
others, sustained JNK activation could mediate the process by a
previous increase in ceramide levels, or not (33, 34, 35),
whereas in cells lacking the MEKK1 gene, in which no JNK activity
exists, apoptosis was enhanced (61). On the other hand,
some data indicate that JNK1 and JNK2 can play different roles
depending on the cell type and/or apoptotic stimuli (36, 62). Our data do not suggest that JNKs are involved in this
process, because although overexpression of JNK2, but not of JNK1,
slightly enhanced basal apoptosis, the expression of a dominant
negative mutant of JNK1 or JNK2, which highly decreased endogenous JNK
activity, was unable to block TNF-induced apoptosis. This can be
related to the fact that TNF did not induce sustained JNK activation
in brown adipocytes. Thus, the activation of JNK was very strong at
10–15 min, but then decreased. The slight increase observed at 2, 5,
and 24 h does not appear to be very important for this apoptotic
process and could be just a consequence of caspase activation. Thus,
other signaling pathways, different from p38MAPK, could be also
involved in this process, although it cannot be ruled out that SB203580
could not completely inhibit TNF-induced apoptosis due to its
toxicity after long-term-treatment (57).
Regarding ERKs, although we found a protective role for ERKs in the
TNF-induced apoptosis in brown adipocytes, as in other cell systems
(21, 22), ERKs do not always play a role in survival
(63), and their actions probably depend on the cell type
and/or the apoptotic signal. Moreover, in our cell system, it is very
likely that other pathways activated by TNF, such as NF-B, may be
involved in survival, as it occurs in other cell types
(47, 48, 49). This would also explain why this process of
apoptosis is not so dramatic, because antiapoptotic pathways attenuate
those proapoptotic pathways that are activated simultaneously. This is
very common for TNF, and in some cases, it is required to block
these antiapoptotic pathways (for example, by inhibition of protein
synthesis) to induce apoptosis (11).
In relation to the roles played by ERKs and p38MAPK in the cell cycle
in brown adipocytes, they have opposite effects as in the process of
apoptosis. Thus, meanwhile, inhibition of the ERK cascade enhances the
TNF-induced inhibition of proliferation (basal and
IGF-I/insulin-induced), and p38MAPK blockade abolishes this effect.
Therefore, ERKs activated by TNF play a positive role in the
regulation of proliferation in brown adipocytes, as it occurs when
activated by IGF-I and insulin (17). In contrast, p38MAPK
mediates the TNF-induced inhibition of cell cycle progression.
Similarly, in fibroblasts, ERKs play a positive role, whereas p38MAPK
has an inhibitory role in cell proliferation (31). Further
evidence of inhibition of cell cycle by p38MAPK has been recently
obtained using an estrogen-induced MEKK3 construct (32).
It seems that most of these effects on cell cycle are correlated with
changes in cyclin D1 expression (31, 32); however, our
data do not indicate any regulation at the level of cyclin D1, whereas
a regulation of PCNA exists. This agrees with the modulation of other
cell cycle-regulated proteins by p38MAPK that has been described
(32). On the other hand, p38s can also activate cell
proliferation (64, 65); it is possible that depending on
the cell type and the stimulus, p38s can have either a positive or a
negative role in cell proliferation.
On the other hand, TNF induces cell death in nonproliferating cells,
and this effect remains unchanged when p38MAPK is inhibited. Therefore,
it seems that inhibition of p38MAPK decreases the number of cells dying
by a process of apoptosis induced by TNF, but the cells that still
die are not proliferating. In contrast, ERK inhibition increases the
number of apoptotic cells compared with TNF, but most of these cells
are positive for PCNA and, hence, they are progressing through the cell
cycle.
In relation to the possible mediators of p38MAPK and ERKs actions,
c-Jun might play a role. Thus, inhibition of ERKs prevents
TNF-induced c-Jun mRNA expression, which would be correlated with
inhibition of proliferation and enhancement of apoptosis. In contrast,
p38MAPK blockade enhances TNF-induced c-Jun mRNA expression, which
could partially account for the reversion of cell cycle arrest and the
reduction of apoptosis. These would be in agreement with the role
proposed for c-Jun as a mediator of proliferation and survival in c-Jun
null mouse embryo fibroblasts (66).
Another possible target for p38MAPK could be NF-B, which plays a
protective role in some processes of apoptosis (47, 48, 49).
It has been described that p38MAPK can increase either its
trans-activation activity (51) or its nuclear
translocation (44) or can even inhibit its activation
(67). However, in brown adipocytes, binding of NF-B to
DNA in response to TNF is not regulated by either p38MAPK or
ERKs.
p38MAPK can also regulate CHOP/GADD153 by phosphorylation
(42) and/or by changes in gene expression
(43). CHOP plays an essential role in apoptosis induced by
endoplasmic reticulum stress (68); it seems also to play a
role in some other processes of apoptosis, such as that induced by Fas
and ceramides (26), as well as in cell cycle arrest
induced by stress (27, 42). However, our data do not
suggest that CHOP is involved in TNF-induced apoptosis and cell
cycle arrest, because the level of this protein is almost undetectable
under these conditions.
CREB or other members of this CREB/ATF family of transcription factors
can be phosphorylated by a mechanism dependent on ERKs
(41), p38ßMAPK (45), phosphoinositol
3-kinase/protein kinase B pathway (69),
ERK/p38/mitogen- and stress-activated protein kinase-1 (MSK1)
pathway (41), protein kinase A, or other kinases,
and in some cases this phosphorylation has been correlated with cell
survival. In brown adipocytes, we have shown here that CREB is already
phosphorylated under control conditions, and no changes are induced in
response to TNF. However, ATF1 is phosphorylated after TNF
treatment by a mechanism dependent on p38MAPK that could be involved in
the regulation of cell cycle and/or apoptosis via p38MAPK. In fact, in
Jurkat cells, anisomycin-induced p38MAPK activation is responsible for
ATF1 phosphorylation and subsequent induction of early growth response
gene-1 (70).
Based on the results presented here, it can be concluded that p38MAPK
plays an important role as a mediator of TNF-induced apoptosis and
cell cycle arrest in brown adipocytes. In contrast, ERKs attenuate the
TNF effects, and JNKs do not seem to play a significant role in
TNF-induced apoptosis. Thus, in fetal brown adipocytes, the
simultaneous activation of ERKs, an antiapoptotic pathway, and
p38MAPK, a proapoptotic pathway, by TNF might be very important for
regulation of the number of cells during the development of BAT in the
perinatal period or even under other circumstances, such as obesity.
The balance between these opposite pathways and some others might
determine the precise number of cells that should die. In addition,
ERKs and p38MAPK play opposite roles in the inhibition of proliferation
and differentiation (our unpublished data) of brown adipocytes.
Thus, ERKs mediate proliferation and survival, and inhibit
differentiation, whereas p38MAPK does the opposite. In this way, TNF
might regulate not only the number of cells in BAT, but also the
differentiation state of these cells. On the other hand, depending on
the presence of other extracellular signals, this balance can be
modified, so the cells would preferentially proliferate, differentiate,
or die.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Recipients of fellowships from the Ministerio de Educación y
Cultura, Spain.
Received April 14, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
causes insulin receptor
substrate-2-mediated insulin resistance and inhibits insulin-induced
adipogenesis in fetal brown adipocytes. Endocrinology 139:1–10
without effects on Pref-1 expression. Endocrinology 138:2776–2783 gene
expression contributes to the antiadipogenic effects of tumor
necrosis-. Mol Endocrinol 10:1457–1466[Abstract]
induces apoptosis in rat fetal brown adipocytes in primary
culture. FEBS Lett 416:324–328[CrossRef][Medline]
induces apoptosis in rat
brown adipocytes. Cell Death Differ 4:771–778
induces apoptosis of human adipose cells. Diabetes 46:1939–1944[Abstract]
B induction by TNF, CD95 and
IL-1. Nature 385:540–544[CrossRef][Medline]
-induced activation of p42mapk/erk2 in
mouse macrophages. J Biol Chem 270:27391–27394-mediated apoptosis
requires Ras and the activation of mitogen-activated protein kinase.
J Biol Chem 271:14560–14566
