Heme Oxygenase-1-derived Carbon Monoxide Requires the Activation
of Transcription Factor NF-
B to Protect Endothelial Cells from Tumor
Necrosis Factor-
-mediated Apoptosis*
Sophie
Brouard
§¶,
Pascal O.
Berberat§
,
Edda
Tobiasch,
Mark P.
Seldon**
,
Fritz H.
Bach§§, and
Miguel
P.
Soares**¶¶
From the Immunobiology Research Center, Department of
Surgery, Beth Israel Deaconess Medical Center, Harvard
Medical School, Boston, Massachusetts, 02115 and the
** Instituto Gulbenkian de Ciência, Apartado 14, 2781-901 Oeiras, Portugal
Received for publication, August 29, 2001, and in revised form, March 1, 2002
 |
ABSTRACT |
We have shown that carbon
monoxide (CO) generated by heme oxygenase-1 (HO-1) protects endothelial
cells (EC) from tumor necrosis 
(TNF-)-mediated apoptosis. This
effect relies on the activation of p38 MAPK. We now demonstrate that
HO-1/CO requires the activation of the transcription factor NF-
B to
exert this anti-apoptotic effect. Our data suggest that EC have basal
levels of NF-B activity that sustain the expression of
NF-B-dependent anti-apoptotic genes required to support
the anti-apoptotic effect of HO-1/CO. Over-expression of the inhibitor
of NF-B (IB) suppresses the anti-apoptotic action of
HO-1/CO. Reconstitution of NF-B activity, by co-expression of
IB with different members of the NF-B family, i.e.
p65/RelA or p65/RelA plus c-Rel, restores the anti-apoptotic effect of
HO-1/CO. Expression of the NF-B family members p65/RelA or p65/RelA
with p50 or c-Rel up-regulates the expression of the anti-apoptotic
genes A1, A20, c-IAP2, and manganese superoxide dismutase (MnSOD). Inhibition of NF-B
activity by over-expression of IB suppresses the expression of
some of these anti-apoptotic genes, i.e.
c-IAP2. Under inhibition of NF-B, co-expression of some
of these anti-apoptotic genes, i.e. c-IAP2 and
A1, restores the anti-apoptotic action of HO-1/CO, whereas expression of A20 or MnSOD cannot. The ability
of c-IAP2 and/or A1 to restore the anti-apoptotic action of HO-1/CO is
abolished when p38 MAPK activation is blocked by over-expression of a
p38 MAPK dominant negative mutant. In conclusion, we demonstrate that HO-1/CO cooperates with NF-B-dependent anti-apoptotic
genes, i.e. c-IAP2 and A1, to
protect EC from TNF--mediated apoptosis. This effect is dependent on
the ability of HO-1/CO to activate the p38 MAPK signal transduction pathway.
| |
INTRODUCTION |
Signaling via "death receptors," such as the tumor necrosis
factor- (TNF-)1
receptor 1 (TNFR-1/CD120a), can trigger endothelial cell (EC) to
undergo apoptosis. Cross-linking of TNFR-1 leads to the recruitment of
intracytoplasmic signal transduction molecules, e.g. TRADD (TNF receptor-associated death domain), FADD (Fas-associated death domain), and RIP (receptor-interacting protein) (1, 2). These
molecules form the death-inducing signaling complex (DISC; reviewed in Refs. 1 and 2), which activates serine proteases, referred
to as caspases (3). Caspase activation by the death-inducing signaling
complex occurs via FADD-dependent recruitment and proximal catalytic cleavage/activation of pro-caspase-8 into the active form of
caspase-8 (4, 5), which activates additional pro-caspases into active
caspases, e.g. caspase-3, that execute the terminal phase of
apoptosis (reviewed in Ref. 3).
Under physiologic conditions, signaling via the TNFR-1 does not lead to
EC apoptosis because TNFR-1 triggers the expression of early responsive
anti-apoptotic genes such as the zinc finger A20 (6), the
bcl-2 family member A1 (7), the
antioxidant manganese superoxide dismutase (MnSOD) (8),
several members of the inhibitor of apoptosis (IAP) family (9),
IEXL-1 (10), and PAI-2 (plasminogen activator
inhibitor type-2) (11). These anti-apoptotic genes prevail over the
pro-apoptotic signals thus preventing TNF--mediated EC apoptosis.
Expression of these anti-apoptotic genes is dependent on the activation
of the transcription factor nuclear factor B (NF-B) (12).
Inhibition of NF-B activity prevents the expression of these
anti-apoptotic genes and thus sensitizes most cell types (13-15),
including EC (12), to undergo TNF--mediated apoptosis.
The NF-B family of transcription factors consists of several homo-
or heterodimeric complexes of the Rel family, i.e.
p50/NF-B1, p65/RelA, c-Rel (Rel), p52/NF-B2, and RelB (reviewed
in Refs. 16 and 17). In quiescent EC, NF-B is thought to be retained in the cytoplasm by a series of inhibitory proteins referred to as
inhibitor of B (IB) (reviewed in Ref. 16). Binding of NF-B to
IB molecules masks the nuclear localization signal in the NF-B
dimers, thereby preventing NF-B nuclear translocation and transcription activity (18). Signaling via TNFR-1 triggers the release
of NF-B dimers from IB molecules via site-specific
phosphorylation, ubiquitination, and subsequent proteolytic IB
degradation through the 26 S proteasome pathway (19, 20). Once released
from IB molecules, NF-B dimers translocate into the nucleus to
bind specific decameric recognition motifs in the promoter region of
NF-B-dependent genes such as the anti-apoptotic genes
A1 (21), A20 (22), MnSOD (23), and
c-IAP2 (24).
Other anti-apoptotic genes are expressed by EC independently of
NF-B. These include heme oxygenase-1 (HO-1) (reviewed in Refs. 25 and 26), a stress-responsive gene encoding a 32-kDa enzyme
that degrades heme into biliverdin, iron, and the gas carbon monoxide
(27). Although present only at basal levels in quiescent EC, HO-1
expression is rapidly up-regulated under oxidative stress conditions
(reviewed in Refs. 25 and 26). We have previously shown that HO-1
protects EC from undergoing apoptosis (28). The three end products
of HO-1 enzymatic activity can potentially act as antioxidants and thus
can exert anti-apoptotic effects (29, 30). However, CO
seems to act in a dominant manner to mediate the anti-apoptotic effect
of HO-1 (31). This effect requires the activation of the p38
mitogen-activated protein kinase (MAPK) signal transduction pathway
(31).
Given that signaling transduction pathways leading to
NF-B activation are critical in modulating EC apoptosis, we
questioned whether HO-1-derived CO cooperates with one or more
NF-B-dependent anti-apoptotic genes to prevent
apoptosis. We found that HO-1 requires NF-B activity and the
expression of A1 or c-IAP2 to exert its anti-apoptotic effect in
EC.
| |
EXPERIMENTAL PROCEDURES |
Cell Cultures--
The murine 2F-2B EC line (American Type
Culture Collection (ATCC), Manassas, VA), primary bovine aortic EC
(BAEC), porcine aortic EC (PAEC), and human umbilical vein EC (HUVEC)
were isolated and cultured as described previously (12, 31, 32).
Expression Plasmids--
The 
-galactosidase expression vector
has been described elsewhere (12). Two vectors encoding rat HO-1
cDNA were used, one under the control of the -actin
enhancer/promoter (
actin/HO-1) (33) and the other under the
control of the cytomegalovirus (CMV) enhancer/promoter
(pcDNA3/HO-1) (31). The murine Bcl-2 expression vector (kind gift
of R. Gerard, University of Texas Southwest Medical Center, Dallas, TX)
has been described elsewhere (34). The porcine IB cDNA
(ECI-6) was expressed in the pcDNA3/HA vector-derived from the
pcDNA3 vector (Invitrogen) as described previously (12). The human
A20 cDNA, cloned into the pAC expression vector was originally
obtained from V. Dixit (Genentech, Inc., South San Francisco, CA) (35)
and was expressed in the HA-tagged pcDNA3 expression vector (kind
gift from C. Ferran in our Immunobiology Research Center) as described
elsewhere (36). Human p65/RelA, c-Rel, and p50 cDNAs were cloned
into the pcDNA3 expression vector (kind gift from Dr. J. Anrather,
Cornell University, New York). The RelA (p65m) DNA binding-deficient
mutant (37) has been described elsewhere (12). P38/CSBP1 MAPK was
amplified from HeLa cDNA by PCR and cloned into the pcDNA3/HA
vector derived from pcDNA3 (Invitrogen) by inserting a DNA fragment
coding for an epitope derived from the hemagglutinin protein of the
human influenza virus hemagglutinin (HA; MYPYDVPDYASL). A dominant
negative mutant of p38/CSBP1 harboring a T180A and a Y182F
substitution, was generated by overlap extension mutagenesis a
described previously (31). This constructed was provided by Dr. J. Anrather (Cornell University, New York) The N-terminally HA-tagged
human A1 expression vector (HA-A1) was provided by Dr. C. Ferran in our
Immunobiology Research Center and has been described elsewhere (38).
The human c-IAP2 cDNA (kind gift from D. Vaux, Australia) was
expressed in the prCMV expression plasmid (kind gift from Dr. D. W.
Ballard Vanderbilt University, Nashville, TN) (39). The human
MnSOD cDNA was obtained from ATCC and cloned (EcoRI)
into the pcDNA3 expression vector (Invitrogen). The human
TNF-R1/pcDNA3 expression vector was obtained from D. V. Goeedel (Tularik, Inc., South San Francisco, CA).
Transient Transfections--
BAEC and 2F-2B EC were transiently
transfected as described elsewhere (12, 31, 40).
-Galactosidase-transfected cells were detected as described
elsewhere (23, 28). Briefly, the number of random fields counted was
determined to have a minimum of 200 viable transfected cells/control
well (without apoptosis-inducing treatment). The number of viable cells
was assessed by evaluating -galactosidase-expressing cells that
retained normal morphology under the apoptosis-inducing treatment,
i.e. TNF- or serum deprivation and control treatment (12,
31, 40). The percent survival was calculated for each DNA
preparation by normalizing the number of viable
-galactosidase-expressing cells counted after the apoptosis-inducing treatment to that counted in the absence of the treatment (100% viability). All experiments were performed in duplicate two to four times.
Recombinant Adenoviruses--
The recombinant -galactosidase
adenovirus was a kind gift of Dr. Robert Gerard (University of Texas
Southwest Medical Center, Dallas, TX). The recombinant IB
adenovirus expressing the porcine IB gene
(ECI-6) has been described elsewhere (41). All
recombinant adenoviruses were produced in 293 cells (ATCC), extracted,
and purified through two cesium chloride gradient ultracentrifugations, and their titer was determined by limiting dilution in 293 cells as
described before (41). Confluent BAEC were infected with a multiplicity
of infection of 200 plaque-forming units/cell as described elsewhere
(31).
Cell Extracts and Western Blot Analysis--
Cell extracts were
prepared and subjected to electrophoresis as described elsewhere (12).
HO-1 was detected using a rabbit antihuman HO-1 polyclonal antibody
(StressGen, Biotechnologies Corp., Victoria, CA). p65/RelA, p50, and
c-Rel were detected using rabbit anti-mouse p50, c-Rel, and RelA
antibodies (Santa Cruz Biotechnology, Santa Cruz, CA).
c-Myc-tagged p65/RelA was detected using an anti-c-Myc
monoclonal antibody (9E10 clone, ATCC). HA-tagged A1, A20, and IB
were detected using a mouse or rat anti-HA monoclonal antibody (Roche
Molecular Biochemicals). c-IAP2-Flag was detected using a
anti-Flag M1 monoclonal antibody (Sigma). Human MnSOD was detected
using a rabbit anti-human MnSOD antibody (Stressgen). -Tubulin was detected using anti-human -tubulin monoclonal
antibody (Roche Molecular Biochemicals). Primary antibodies were
detected using horseradish peroxidase-conjugated donkey anti-rabbit or goat anti-mouse IgG secondary antibodies (Pierce). Peroxidase enzymatic
activity was visualized using the Enhanced Chemiluminescence assay
(Amersham Biosciences), according to the manufacturer's instructions
and stored in the form of photoradiographs (BiomaxTMMS,
Eastman Kodak, Rochester, NY).
Cell Treatment and Reagents--
Water soluble Actinomycin-D
(Act.D, Sigma) was dissolved in phosphate-buffered saline and
added to the culture medium, 24 h after EC transfection. The
concentration of Act.D required to sensitize EC to TNF--mediated
apoptosis was 10 µg/ml for 2F-2B EC and 0.1 µg/ml for BAEC. Human
recombinant TNF- (R&D Systems, Minneapolis, MN) was dissolved in
phosphate-buffered saline, 0.1% bovine serum albumin and added to the
culture medium (10 ng/ml, 50 units/ml) 24 h after EC transfection.
When used in combination with Act.D, EC were exposed to TNF- for a
period of 8 h. When used in the absence of Act.D, EC
(i.e. overexpressing IB) were exposed to TNF- for a
period of 16 h. Serum deprivation was carried out by exposing EC
to 1% fetal calf serum during 30 h.
CO Exposure--
Briefly, CO at a concentration of 1% (10,000 ppm) in compressed air was mixed with balanced air (21%
O2) in a stainless steel mixing cylinder before entering
the exposure chamber. CO concentrations were controlled by varying the
flow rates of CO in a mixing cylinder before delivering it to the
chamber. Because the flow rate is primarily determined by the
O2 flow, only the CO flow was changed to deliver the final
concentration to the exposure chamber. A CO analyzer (Interscan Corp.,
Chatsworth, CA) was used to measure CO levels in the chamber. Cells
were exposed to CO for 1 h before stimulation with TNF- or
serum deprivation and continuously thereafter.
Reporter Assays--
Cellular extracts were assayed for
-galactosidase activity using the Galacto-Light protocol (Applied
Biosystems, Tropix Inc., Bedford, MA). Luciferase activity was assayed
by adding 10 µl of cellular extract to 90 µl of a solution
containing 24 mM glycl-glycine (pH 7.8), 2 mM
ATP (pH 7.5), and 10 mM MgSO4. Samples were
read on the Microlumat LB 96P luminometer (EG&G Berthold, Wildbad, Germany) using an injection mix consisting of 24 mM
glycl-glycine and 0.1 mM luciferin (Sigma). Luciferase
activity was normalized for -galactosidase as follows: luciferase
activity/-galactosidase activity × 1000. Normalized luciferase
activity is shown in arbitrary units.
Electrophoretic Mobility Shift Assays (EMSA)--
Nuclear
extracts were prepared as described elsewhere (12, 42). All buffers
were supplemented with 0.1 mM
L-1-tosylamido-2-phenylethyl chloromethyl ketone, 0.1 mM 1-chloro-3-tosylamido-7-amino-2-heptanone, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM
dithiothreitol, 1 µg/ml aprotinin, and 1 µg/ml leupeptin. Equal
amounts of nuclear extracts (5 µg) were incubated (30 min at room
temperature) with 100,000 cpm of double-stranded,
[
-32P]ATP-radiolabeled NF-B consensus
oligonucleotide 5'-AGT TGA GGG GAC TTT CCC AGG C-3' (Promega), and the
resulting DNA-protein complexes were separated on a 6% polyacrylamide
gel in Tris-glycine-EDTA buffer at pH 8.5. The mass of the probes was
~20 fmol. Supershift assays were carried out in the same manner, but
nuclear extracts were incubated with 2 µg of anti-human p65/RelA
N-terminal antibody (Santa Cruz Biotechnology) for an additional hour
before electrophoresis.
RT-PCR--
RNA was extracted using TRIzol (Invitrogen) and
reverse-transcribed into cDNA with the RNA PCR Kit (TaKaRa,
PanVera, Madison, WI). A total of 2 µl of cDNA was amplified in a
50-µl reaction mix containing 10 mM dNTPs, 50 pg of
5'-prime and 3'-prime oligonucleotides, 2.5 units of
LA-Taq polymerase (TaKaRa) and MgCl2, specific
to each primer pair used. The primers for murine and human
A20 (620 bp) (5'-AAT ATG CGG AAA GCT GTG AAG, 3'-GAT TCC AAA
CTT CTT AGC ATT), A1 (257 bp) (5'-AAA GAA TCT GAAGTC AT,
3'-ATA GGT AAG AGG ACA C), bcl-xL (540 bp) (5'-GCC
AGT GAG CTT CCC GTT CAG C, 3'-CAG AGC AAC CGG GAG CTG GT),
MnSOD (511 bp) (5'-AAC GCG CAG ATC ATG CAG CTG C, 3'-ACA TTC
TCC CAG TTG ATT CAC T), c-IAP2 (567 bp) (5'-TGG GCT TCA GTA
GGA GCC TGG T, 3'-ACT ACT AGA TGA CCA CAC GGA A), E-selectin (1157 bp)
(5'-GGA TTG GAA TCA GAA AAG TCA A, 3'-GGA CTT GTA GGT GAA TTC TCC A),
-actin (525 bp) (5'-GCC ATC CTG CGT CTG GAC CTG G, 3'-TAC TCC TGC
TTG CTG ATC CAC A), and IB (942 bp) (5'-TGG ACG ACC GCC ACG ACA
GCG GC, 3'-CAG TCG ACC GGG TCG ACG ACG ACA TAG GCC CA) were obtained
from Invitrogen). PCR reactions were performed after a 4-min
denaturation at 94 °C a repeating the cycle 94 °C, 55 °C, and
72 °C each for 1 min for the number of cycles specific for each
primer pair in a Peltier Thermal Cycler PTC-200 (MJ Research,
Las Vegas, NV). PCR products (10-20 µl) were analyzed in an ethidium
bromide-stained 1% agarose gel.
Immunocytochemistry--
2F-2B EC were cultured on gelatinized
glass slides (PerkinElmer Life Sciences), fixed in 75% acetone, and
stained with anti-paxillin antibody (Upstate Biotechnology, Lake
Placid, NY), anti-N-terminal p65/RelA antibody (2 µg/ml, Santa
Cruz Biotechnology, SC 372), or anti-p65/RelA nuclear localization
domain sequence-specific antibody (Roche Molecular Biochemicals).
Primary antibodies were detected using biotinylated secondary
antibodies and biotinylated-horseradish peroxidase-coupled streptavidin
reaction (Pierce). Nonspecific purified Ig isotype was used as negative controls.
Confocal Microscopy--
BAEC were cultured on gelatinized glass
slides (PerkinElmer Life Sciences), fixed in 3.7% paraformaldehyde
(Sigma), and stained with an anti-N-terminal p65/RelA antibody (2 µg/ml, Santa Cruz Biotechnology, SC 372). Primary antibody was
detected using fluorescein isothiocyanate-labeled goat anti-rabbit
antibody (F-9262, Pierce). Fluorescent labeling was detected
(
ex = 488 nm; em = 518 nm) using a
multiphoton confocal microscope (BioRad, MRC 1024) equipped with
LaserSharp, version 3.2 software (BioRad).
| |
RESULTS |
HO-1-derived CO Requires NF-B Transcriptional Activity to
Suppress EC Apoptosis--
Inhibition of transcription by Act.D
sensitized control EC transfected with pcDNA3 to undergo
TNF--mediated apoptosis (60-70% apoptosis) (Fig.
1A). Over-expression of HO-1
or Bcl-2-protected EC from apoptosis (5-10% apoptosis) (Fig.
1A) (12, 31). When EC over-expressed IB, a gene that
suppresses NF-B activation (41), HO-1 was no longer able to prevent
TNF--mediated EC apoptosis (60-70% apoptosis) (Fig.
1A). In contrast, Bcl-2 protected EC from TNF--mediated
apoptosis when IB was over-expressed (5-15% apoptosis) (Fig.
1A). As previously shown (12, 31) exogenous CO protected EC
cells from TNF--mediated apoptosis (Fig. 1B). Exogenous
CO also protected EC from serum deprivation-induced apoptosis (Fig.
1C). CO no longer protected EC from TNF--mediated apoptosis when IB was over-expressed (50-60% apoptosis) (Fig. 1, A and B). However, CO was still able to
protect EC from serum deprivation-mediated apoptosis when IB was
over-expressed (5-10% apoptosis) (Fig. 1C). These data
indicate that CO requires the activation of NF-B to protect EC from
TNF- but not from serum deprivation-mediated apoptosis.
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 1.
HO-1-derived CO requires
NF-B transcriptional activity to suppress
endothelial cell apoptosis. A, BAEC were transiently
transfected with -galactosidase (300 ng/3 × 105
cells) and when indicated (+) with HO-1 (-actin/HO-1; 700 ng/3 × 105 cells), bcl-2 (700 ng/3 × 105 cells), and/or IB (500 ng/3 × 105 cells) expression vectors. Gray and
black histograms represent EC treated with Act.D or Act.D
plus TNF-, respectively. Act.D (10 µg/ml) and TNF- (50 units/ml) were added to the culture medium 24 h after transfection
as described under "Experimental Procedures." Percent survival of
transfected EC was calculated 8 h after the addition of Act.D or
Act.D plus TNF- as described under "Experimental Procedures."
Results shown are the mean ± standard deviation from duplicate
wells in two independent experiments (n = 4). The
percentage of cell survival in HO-1-transfected EC was increased in a
very significant manner as compared with control EC transfected with
pcDNA3 (p = 0.0038, unpaired t test).
There was no significant increase in the percentage of survival of EC
co-transfected with HO-1 and IB versus EC transfected
with pcDNA3 (p = 0.4970, unpaired t
test) or IB (p = 0.4759, unpaired t
test) B, mouse 2F-2B EC were co-transfected with IB
(500 ng/3 × 105 cells) plus -galactosidase (300 ng/3 × 105 cells) expression vectors and when
indicated (+) were exposed to exogenous CO (10,000 ppm), as described
under "Experimental Procedures." Gray and black
histograms represent EC treated with Act.D and Act.D plus TNF-,
respectively, as described in A, percent survival was
calculated as in A. Results shown are the mean ± standard deviation from duplicate wells from one of three similar
experiments. C, mouse 2F-2B EC were transfected as described
in B. When indicated (+) EC where exposed to exogenous CO
(10,000 ppm) as described under "Experimental Procedures."
Twenty-four hours after transfection, EC were exposed to either 10%
(gray histograms) or 1% (black histograms) fetal
calf serum for 30 h. Percent survival was calculated as in
A and B. Results shown are the mean ± standard deviation from duplicate wells in two independent experiments
(n = 4). Percent of cell survival increased in a highly
significant manner in EC exposed to CO (+) versus EC not
exposed to CO () whether or not EC were transfected with IB
(p < 0.0001, unpaired t test).
D, BAEC were transiently transfected with a
NF-B-dependent luciferase (300 ng/3 × 105 cells) and -galactosidase (300 ng/3 × 105 cells) reporters plus or minus IB (500 ng/3 × 105 cells). NF-B activity was induced by transient
over-expression of a p65/RelA (10, 102 or
103ng/3 × 105 cells), as described before
(12). Results shown are mean ± standard deviation from one of
three similar experiments and are expressed in arbitrary luciferase
units (A.U.) normalized for -galactosidase expression.
E, mouse 2F-2B EC were transiently co-transfected with
-galactosidase (300 ng/3 × 105 cells). When
indicated (+) EC were co-transfected HO-1 (-actin/HO-1, 700 ng/3 × 105 cells), IB (500 ng/3 × 105 cells), c-Myc-tagged p65/RelA (1000 ng/3 × 105 cells) and/or a c-Myc-tagged DNA binding deficient
mutant of p65/RelAm (1000 ng/3 × 105
cells) expression vectors. Gray and black
histograms represent EC exposed to medium or TNF- (24 h after
transfection; 50 units/ml, 16 h), respectively. Percent survival
of transfected EC was evaluated as in A-C. Results shown
are the mean ± standard deviation from duplicate wells in two
independent experiments (n = 4). Percent of cell
survival was very significantly increased in IB plus p65
versus IB transfected EC (p = 0.0036, unpaired t test). F, endogenous and
over-expressed IB was detected by Western blot using an
anti-human N-terminal IB polyclonal antibody (C21, Santa Cruz).
Over expressed c-Myc-tagged p65/RelA and p65/RelAm were
detected using an anti-c-Myc monoclonal antibody.
|
|
The observation that IB abrogates the ability of HO-1 and/or CO
to prevent TNF--mediated apoptosis (Fig. 1, A and
B) could be attributed to an intrinsic pro-apoptotic effect
of IB that would act independently of its ability to suppress
NF-B activity. To test this hypothesis, we analyzed whether
reconstitution of NF-B activity, under IB over-expression,
would restore the anti-apoptotic effect of HO-1. Over-expression of
p65/RelA resulted in high levels of NF-B transcriptional activity in
EC (Fig. 1D) in a dose-dependent manner, in that
increasing amounts of p65/relA resulted in increasing levels of NF-B
transcriptional activity (Fig. 1D). Over-expression of
IB inhibited p65/RelA transcriptional activity only when p65/RelA
was expressed at low levels (10 ng/3 × 105 cells)
(Fig. 1D). When the amount of p65/RelA was increased to levels higher then 100 ng/3 × 105 cells, IB
over-expression was no longer fully able to suppress p65/RelA
transcriptional activity (Fig. 1D), a result consistent with
our previous observations (12). Over-expression of IB sensitized
EC to TNF--mediated apoptosis in the absence of Act.D (65-75%
apoptotic EC) (Fig. 1E) (12). Co-expression of p65RelA with
HO-1 restored the anti-apoptotic effect of HO-1 and prevented TNF--mediated apoptosis of IB-expressing EC (2-5% apoptotic EC) (Fig. 1E). This protective effect was not observed when
EC were co-transfected with HO-1 plus a DNA binding-deficient mutant of
p65/RelA, which has no transcriptional activity (37) (Fig. 1E). Expression of HO-1, IB, and p65/RelA proteins was
confirmed by Western blot (Fig. 1F).
Over-expression of HO-1 or Exposure to Exogenous CO Does Not
Activate NF-B in EC--
Given that HO-1/CO requires NF-B
activity to exert its anti-apoptotic effect, we tested whether HO-1
induced NF-B nuclear translocation/activity in EC. Transient HO-1
over-expression in EC did not induce a detectable increase in NF-B
nuclear translocation and/or NF-B binding to NF-B-specific DNA
binding consensus sequences as compared with quiescent EC (Fig.
2A). In a similar manner, over-expression of HO-1 was not associated with detectable increase in
NF-B transcriptional activity (Fig. 2B). Over-expression
of HO-1 protein was confirmed by Western blot (Fig. 2C).
Exposure of EC to exogenous CO (10,000 ppm, 2 h) did not
cause an significant increase in the nuclear translocation of NF-B
(p65/RelA) as analyzed by confocal microscopy (Fig. 2D) and
EMSA (data not shown).
View larger version (55K):
[in this window]
[in a new window]
|
Fig. 2.
Transient over-expression of HO-1
does not modulate NF-B activation.
A, specific binding of NF-B dimers to the NF-B
consensus DNA binding motif, 5'-AGT TGA GGG GAC TTT CCC AGG C-3', was
detected by EMSA using nuclear extracts from nontransfected
(NT), control (pcDNA3), or
HO-1-transfected 2F-2B EC. Values in HO-1-transfected extracts
indicate the amount of plasmid used/3 × 105 cells. EC
were either not stimulated () or stimulated with TNF- (+) (10 ng/ml; 50 units/ml) 20 min before nuclear extraction. Specificity of
the assay was tested using free radiolabeled consensus oligonucleotide
in absence of nuclear extracts (lane I) as well as by DNA
binding competition with 50 ng of nonradiolabeled oligonucleotide
(lane II) or by competition with 50 ng of non-radiolabeled
irrelevant oligonucleotide (lane III). Competition controls
were carried out using nuclear extracts from TNF- stimulated (50 units/ml, 30 min) EC. Identification of the different NF-B
dimeric proteins was carried out using an antibody specific for
p65/RelA as described under "Experimental Procedures."
B, BAEC were transiently transfected with a
NF-B-dependent luciferase (300 ng/3 × 105 cells) plus a CMV-driven -galactosidase (300 ng/3 × 105 cells) reporter. When indicated, EC were
co-transfected with TNFR-1 (pcDNA3 + TNFR-1)
or HO-1 (pcDNA3 + HO-1) expression vectors.
Untreated and TNF- (50 units/ml, 6 h)-treated control EC
(pcDNA3) were used as negative and positive controls, respectively.
Results shown are the mean ± standard deviation from duplicate
wells taken from one of three similar experiments and are expressed in
arbitrary luciferase units (A.U.) normalized for
-galactosidase expression. C, HO-1 expression was
detected by Western blot using a polyclonal rabbit antibody directed
against human HO-1. Values indicated the amount of pcDNA3/HO-1
expression vector used per well. D, BAEC were untreated
(control), treated with LPS (1 µg/ml; 20 min), or exposed to
exogenous CO (10,000 ppm, 2 h) and stained with an anti-p65/RelA
antibody as described under "Experimental Procedures." Fluorescent
staining was analyzed by confocal microscopy as described under
"Experimental Procedures." Notice the cytoplasmic and nuclear
staining (arrows). All magnification is 100×.
|
|
Quiescent EC Have Basal Levels of Nuclear NF-B--
The
data illustrated in Fig. 3 suggest that
quiescent BAEC have basal levels of nuclear NF-B (p65/RelA),
evidenced by confocal microscopy using an antibody that recognizes the
N-terminal region of p65/RelA (Fig. 3). Nuclear p65/RelA was also seen
in quiescent 2F-2B EC by immunocytochemistry using antibodies that
recognizes either the N-terminal region of p65/RelA or the nuclear
localization signal region of p65/RelA, which is rendered
accessible when p65/RelA is released from IB molecules (Fig.
4A). The presence of nuclear p65/RelA in quiescent EC (Figs. 3 and 4A) correlated with
the detection of p65/RelA containing NF-B dimers that bound to
NF-B-specific DNA binding sequences in EMSA (Fig. 4B).
Quiescent 2F-2B EC had two types of nuclear NF-B dimers containing
p65/RelA (referred to as dimers 2 and 3; Fig.
4B). Stimulation of 2F-2B EC with lipopolysaccharide (LPS)
resulted in increased nuclear translocation of dimers 2 and 3 as well
as in nuclear translocation of an additional NF-B dimer containing
p65/RelA (referred to as dimer 4; Fig. 4B).
NF-B dimers 1 and 2 were also detected in quiescent primary PAEC
(Fig. 4C) and BAEC, which had an additional p65/RelA
containing NF-B dimer (referred to as dimer 5; Fig. 4D).
Over-expression of IB in BAEC suppressed DNA binding of dimer 2 and significantly decreased that of dimers 3-5 (Fig.
4D).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 3.
Quiescent EC have basal levels of nuclear
NF-B dimers containing p65/relA. BAEC
were either untreated (A and B) or treated with
LPS (C) (1 µg/ml; 20 min) and stained with rabbit IgG
(A) or an anti-p65/RelA rabbit IgG (B and
C) as described under "Experimental Procedures."
Fluorescent staining was analyzed by confocal microscopy as described
under "Experimental Procedures." Notice cytoplasmic and nuclear
localization (white arrows) of p65/RelA in quiescent BAEC.
All magnification is 100×.
|
|
View larger version (56K):
[in this window]
[in a new window]
|
Fig. 4.
Quiescent EC have basal levels of
free nuclear p65/RelA. A, untreated 2F-2B EC were
stained with anti-paxillin antibody and counter-stained with
hematoxylin (a). Untreated (b) or TNF- treated
(10 µg/ml, 30 min) (c) 2F-2B EC were stained with an
anti-p65/RelA purified rabbit antibody (IgG) and counter-stained with
hematoxylin. Untreated 2F-2B EC were stained with a nonspecific rabbit
IgG and counter-stained with hematoxylin (d). Untreated
(e) or TNF- treated (10 µg/ml; 30 min) (f)
2F-2B EC were stained with an anti-p65/RelA rabbit purified antibody
(IgG) without hematoxylin counter-staining. Untreated (g) or
TNF--treated (10 µg/ml, 30 min) (h) 2F-2B EC were
stained with an anti-p65/RelA antibody recognizing the nuclear
localization sequence exposed in non-IB-bound p65/RelA without
hematoxylin counter-staining. Notice the cytoplasmic (Cyt.)
and nuclear (Nuc.) staining (black arrows). All
magnification is 80×. B, specific binding of NF-B dimers
to the consensus DNA binding motif, 5'-AGT TGA GGG GAC TTT CCC AGG
C-3', was detected by EMSA using nuclear extracts from
untreated (Medium) or LPS (5 µg/ml, 30 min) 2F-2B.
Specificity of the assay for the consensus analyzed was tested using
free radiolabeled consensus oligonucleotide in the absence of nuclear
extracts () as well as by DNA binding competition with 50 ng of
nonradiolabeled oligonucleotide (CP) and by competition with
50 ng of nonradiolabeled irrelevant oligonucleotide (CPm).
Identification of the different NF-B heterodimeric proteins
was carried out using an antibody specific for the N-terminal region of
human p65 (-p65), as described under "Experimental Procedures."
C, specific binding of NF-B dimers to the consensus DNA
binding motif, 5'-AGT TGA GGG GAC TTT CCC AGG C-3', was detected by
EMSA as described in B using nuclear extracts from untreated
(Medium) PAEC. D, specific binding of NF-B
dimers to the consensus DNA binding motif, 5'-AGT TGA GGG GAC TTT CCC
AGG C-3', was detected by EMSA, as described in B and
C, using nuclear extracts from untreated (Medium)
or LPS (5 µg/ml; 30 min)-treated BAEC. The specificity of the assay
for the consensus analyzed was tested as described in B
using nuclear extracts from LPS-treated BAEC. Identification of the
different NF-B heterodimeric proteins was carried out as described
in B. When indicated BAEC were infected with a
-galactosidase (gal.) or an IB recombinant
adenoviruses as described under "Experimental Procedures."
NI indicates noninfected EC; *, indicates nonspecific
labeling; 1, indicates p65/RelA-containing dimer bound to
anti-p65 antibody. Based on their p65/RelA content and relative
molecular weights, we predict that dimers denominated as 2,
3, 4, and 5 correspond to p65/RelA (65 kDa) + c-Rel (75 kDa), RelB (68kDa), and p52 and p50 (49 kDa),
respectively.
|
|
NF-B Activity Restores the Anti-apoptotic Effect of HO-1-derived
CO--
We tested whether basal level of NF-B activity would
support the anti-apoptotic effect of HO-1-derived CO. Transient
over-expression of p65/RelA, p65/RelA plus p50, or p65/RelA plus c-Rel
induced NF-B transcriptional activity in EC (Fig.
5A). p65/RelA was more efficient in doing so than p65/RelA plus p50 or c-Rel (Fig.
5A). Co-expression of IB with these Rel family members
inhibited NF-B transcriptional activity (Fig. 5A).
However, NF-B activity was still 2-3 times higher in EC that
co-expressed IB with the different Rel proteins compared with EC
that expressed IB alone (Fig. 5A). When co-expressed
with IB at these levels, p65/RelA or p65/RelA plus p50 or c-Rel
per se did not protect EC from TNF--mediated apoptosis
(60-70% of EC apoptosis) (Fig. 5B). However, at the same
level of expression both p65/RelA and p65/RelA plus c-Rel restored the
anti-apoptotic effect of CO in EC that over-expressed IB (Fig.
5B). Co-expression of p65/RelA with p50 did not restore the
anti-apoptotic effect of CO (Fig. 5B), which was not
correlated with a lower transcriptional activity of p65/RelA plus p50
versus p65/RelA plus c-Rel (Fig. 5A). Similarly,
when EC were co-transfected with HO-1 and IB, p65/RelA or
p65/RelA plus c-Rel with IB restored the anti-apoptotic effect of
HO-1 (data not shown). Expression of p65/RelA, p50, and c-Rel was
detected by Western blot using antibodies that recognize both the
endogenous and over-expressed forms of these NF-B family members
(Fig. 5C). Over-expressed IB was detected by Western
blot using an anti-HA antibody (Fig. 5C).
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 5.
NF-B activation is
required to sustain the anti-apoptotic effect of HO-1-derived CO.
A, BAEC were transiently transfected with an
NF-B-dependent luciferase (300 ng/3 × 105 cells) plus a CMV-driven -galactosidase (300 ng/3 × 105 cells) reporter. When indicated (+) EC
were transiently co-transfected with p65/RelA (200 ng/3 × 105 cells), p65/RelA (100 ng/3 × 105
cells) plus p50 (100 ng/3 × 105 cells), or p65/RelA
(100 ng/3 × 105 cells) plus c-Rel (100 ng/3 × 105 cells) with (+) or without () IB (300 ng/3 × 105 cells). Results shown are the mean ± standard
deviation from duplicate wells in one representative experiment of
three similar ones and are expressed in arbitrary luciferase units
(A.U.) normalized for -galactosidase expression.
B, 2F-2B EC were transfected as described in A,
without the NF-B-dependent reporter. When indicated (+)
EC were exposed to exogenous CO (10,000 ppm) as described under
"Experimental Procedures." Gray and black
histograms represent EC exposed to medium or TNF- (24 h after
transfection; 50 units/ml, 16 h), respectively. Percent survival
of transfected EC was calculated as described under "Experimental
Procedures." Results shown are the mean ± standard deviation
from duplicate wells in three independent experiments
(n = 6). Percent of cell survival was increased in a
highly significant manner in IB plus p65 versus
IB-transfected EC exposed to CO (p = 0.0001, unpaired t test). There was no significant increase in the
percentage of survival of IB plus p65/p50 versus
IB-transfected EC exposed to CO (p = 0.4752, unpaired t test). Percent of cell survival was increased in
a highly significant manner in IB plus p65/c-Rel
versus IB-transfected EC exposed to CO
(p = 0.0002, unpaired t test). C,
transfected HA-tagged IB was detected by Western blot using an
anti-HA epitope monoclonal antibody as described under "Experimental
Procedures." p65/RelA, p50, and c-Rel were detected using polyclonal
antibodies directed against the human sequence of these NF-B family
members as described under "Experimental Procedures."
|
|
Quiescent EC Have Basal Expression of NF-B-dependent
Anti-apoptotic Genes--
We hypothesized that the anti-apoptotic
effect associated with basal NF-B activity acted via the expression
of NF-B-dependent anti-apoptotic genes. Consistent with
this hypothesis, quiescent EC expressed basal levels of mRNA
encoding the NF-B-dependent anti-apoptotic genes
A1, A20, MnSOD, and c-IAP2 as detected
by RT-PCR (Fig. 6, A and
D). These cells did not express mRNA encoding the
pro-inflammatory gene E-selectin, illustrating their "quiescent" state (Fig. 6A). The following set of observations
demonstrated that expression of A1, A20, MnSOD, and
c-IAP2 is NF-B-dependent in these cells: i)
up-regulation of these genes by TNF- was abrogated when NF-B
activation was blocked by IB over-expression (data not shown);
ii) over-expression of p65/RelA, p65/RelA with p50, or p65/RelA with
c-Rel up-regulated the expression of these genes but not that of the
non-NF-B-dependent anti-apoptotic gene
bcl-xL (Fig. 6, B and C); and
iii) over-expression of IB suppressed the basal level of
expression of some of these genes, i.e. c-IAP2 but not A1, A20, or MnSOD (Fig. 6D).
Up-regulation of the expression of these anti-apoptotic genes
correlated with the ability of p65/RelA, p65/RelA plus p50 or p65/RelA
plus c-Rel to protect EC from TNF- plus Act.D-mediated apoptosis
(Fig. 6E). That the anti-apoptotic effect of these Rel
family members is mediated through the up-regulation of A1, A20, MnSOD,
and c-IAP2 is supported by the observation that over-expression of
these anti-apoptotic genes was sufficient per se to protect
EC from TNF- plus Act.D-mediated apoptosis (Fig. 6F).
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 6.
A1, A20, MnSOD, and
c-IAP2 are
NF-B-dependent anti-apoptotic
genes that protect EC from TNF--mediated
apoptosis. A, HUVEC were either untreated () or
treated (+) with TNF- (50 units/ml, 2 h) before RNA extraction.
Expression of A1, A20, MnSOD,
c-IAP2, E-selectin, and -actin mRNA was detected by
RT-PCR as described under "Experimental Procedures." Notice that in
the absence of TNF- stimulation, HUVEC express basal levels of
A1, A20, MnSOD, and c-IAP2
but not E-selectin. B, 2F-2B EC were transiently transfected
with p65/RelA (200 ng/3 × 105 cells), p65/RelA (100 ng/3 × 105 cells) plus p50 (100 ng/3 × 105 cells), or p65/RelA (100 ng/3 × 105
cells) plus c-Rel (100 ng/3 × 105 cells). Control EC
were transfected with pcDNA3. Expression of different
anti-apoptotic genes was analyzed by RT-PCR as described under
"Experimental Procedures." C, the results shown in
B were normalized to -actin expression and expressed as
fold increase as compared with control EC transfected with pcDNA3.
D, HUVEC were either not infected or infected with
-galactosidase (-gal.) or IB recombinant
adenoviruses as described under "Experimental Procedures." DNA was
extracted 24 h after adenoviral infections, and the expression of
anti-apoptotic genes was analyzed by RT-PCR as described in
A. Notice that IB over-expression suppressed the
expression of c-IAP2 but not that of the other
anti-apoptotic genes analyzed. The results shown are representative of
two independent experiments. E, 2F-2B EC were transiently
transfected with a -galactosidase reporter (300 ng/3 × 105 cells) plus p65/RelA (1000 ng/3 × 105
cells), p65/RelA (500 ng/3 × 105 cells) plus p50 (500 ng/3 × 105 cells), or p65/RelA (500 ng/3 × 105 cells) plus c-Rel (500 ng/3 × 105
cells). Gray and black histograms represent EC
exposed to Act.D (10 µg/ml) or Act.D (10 µg/ml) plus TNF- (50 units/ml) at 24 h after transfection for a period of 8 h.
Percent survival of transfected EC was calculated as described under
"Experimental Procedures." The results shown are the mean ± standard deviation from duplicate wells in four independent experiments
(n = 8). Notice the highly significant increase in the
percent survival of EC transfected with p65/RelA, p65/RelA plus p50, or
p65/RelA plus p50 versus control EC transfected with
pcDNA3 (p < 0.0001, unpaired t test).
F, 2F-2B EC were transiently transfected with a
-galactosidase reporter (300 ng/3 × 105 cells) and
A1, A20, MnSOD, or c-IAP2
(1000 ng/3 × 105 cells). Gray and
black histograms represent EC exposed to Act.D (10 µg/ml)
or Act.D plus TNF- (50 units/ml) at 24 h after transfection for
a period of 8 h. Percent survival of transfected EC was calculated
as described under "Experimental Procedures." The results shown are
the mean ± standard deviation from duplicate wells in three
independent experiments (n = 6). Notice the significant
increase in the percent survival of EC transfected with A1
(p = 0.0006, unpaired t test), A20
(p = 0.0022, unpaired t test), MnSOD
(p = 0.0004, unpaired t test), and c-IAP2
(p = 0.019, unpaired t test)
versus control EC transfected with pcDNA3.
|
|
The Anti-apoptotic Effect of HO-1 Requires the Expression of
NF-B-dependent Anti-apoptotic Genes--
To mimic basal
level of expression of A1, A20, MnSOD, and c-IAP2
in quiescent EC the expression/function of these genes was analyzed in
a dose-responsive manner under inhibition of endogenous NF-B
activity by IB. Co-expression of A1 or A20 with IB
protected EC from TNF--mediated apoptosis (Fig.
7, A and B). This
effect was dose-dependent in that higher levels of A1 or
A20 expression increased protection (Fig. 7, A and
B). Co-expression of MnSOD with IB did not prevent EC
apoptosis, a situation that mimics that of HO-1 (Fig. 7C).
Co-expression of c-IAP2 with IB also protected EC from apoptosis
(Fig. 7D). However, the protective effect of c-IAP2 was not
strictly dose-dependent in that protection was lost when
c-IAP2 was expressed above a certain threshold level (Fig.
7D). This effect was not altered when TNF
receptor-associated receptor 2 was co-expressed with c-IAP2
(data not shown), a phenomenon reported to occur in other cell types
(43). We tested whether expression of these anti-apoptotic genes at a
level that, per se, would not prevent EC from undergoing
apoptosis (i.e. attempting to mimic the situation found
in quiescent EC) would reconstitute the anti-apoptotic effect of HO-1.
When co-expressed with IB, A1 (250 ng/3 × 105
cells), A20 (250 ng/3 × 105 cells), MnSOD (25 ng/3 × 105 cells), or c-IAP2 (25 ng/3 × 105 cells) did not prevent EC apoptosis (Fig.
8A). However, when co-expressed with IB, A1 and c-IAP2 but not A20 or MnSOD restored the anti-apoptotic effect of HO-1 (Fig. 8A). We tested
whether CO would act in a similar manner to protect EC from
TNF--mediated apoptosis. Co-expression of suboptimal levels of A1 or
c-IAP2 with IB supported the anti-apoptotic function of exogenous
CO (10,000 ppm), but A20 and MnSOD did not (Fig. 8B).
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 7.
Titration of the anti-apoptotic effect of
NF-B-dependent anti-apoptotic
genes in EC. 2F-2B EC were transiently transfected with a
-galactosidase reporter (300 ng/3 × 105 cells),
IB (500 ng/3 × 105 cells), A1
(A), A20 (B), MnSOD
(C), or c-IAP2 (D) expression vectors
used at the amounts indicated. Control EC were transfected with
-galactosidase plus pcDNA3 expression vectors. Apoptosis was
induced by TNF- (24 h after transfection; 50 units/ml, 16 h).
Percent survival of transfected EC was calculated as described under
"Experimental Procedures." The results shown are mean ± standard deviation from duplicate wells in one of three similar
experiments. Expression of HA-tagged A1, HA-tagged A20, MnSOD, and
Flag-tagged c-IAP2 was detected in BAEC by Western blot using a rat
anti-HA, mouse anti-Flag, or rabbit anti-MnSOD antibodies as described
under "Experimental Procedures." Notice that transfected
human MnSOD (hMnSOD, 23 kDa) can be distinguished from
endogenous BAEC (bMnSOD, 19 kDa) based on the higher
molecular mass of the human form of MnSOD (23 kDa).
|
|
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 8.
Functional interaction between
NF-B-dependent anti-apoptotic
genes and HO-1-derived CO in protecting EC from apoptosis.
A, 2F-2B EC were transiently transfected with a
-galactosidase reporter (300 ng/3 × 105 cells) and
when indicated (+) with IB (500 ng/3 × 105
cells), HO-1 (700 ng/3 × 105 cells) and/or A1 (250 ng/3 × 105 cells), A20 (250 ng/3 × 105 cells), MnSOD (50 ng/3 × 105 cells),
or c-IAP2 (50 ng/3 × 105 cells) expression vectors.
A1, A20, MnSOD, and c-IAP2 were used at suboptimal amounts, which
per se do not prevent EC apoptosis (see Fig. 7).
Gray and black histograms represent EC exposed to
medium or TNF- (24 h after transfection; 50 units/ml; 16 h),
respectively. Percent survival of transfected EC was calculated as
described under "Experimental Procedures." The results shown are
the mean ± standard deviation from duplicate wells in three
independent experiments (n = 6). Notice the highly
significant increase in the percent survival of EC transfected with
HO-1 plus A1 (p = 0.0008, unpaired t test)
or c-IAP2 (p = 0.0002, unpaired t test)
versus EC that do not express HO-1 but express A1 or c-IAP2,
respectively. B, 2F-2B EC were transiently transfected with
a CMV-driven -galactosidase reporter (300 ng/3 × 105 cells) and when indicated (+) with IB (500 ng/3 × 105 cells) and/or A1 (250 ng/3 × 105 cells), or c-IAP2 (50 ng/3 × 105
cells) as described in A. When indicated EC were exposed to
CO (10,000 ppm) as described under "Experimental Procedures."
Gray and black histograms represent EC exposed to
medium or TNF- (24 h after transfection; 50 units/ml, 16 h),
respectively. Percent survival of transfected EC was calculated as
described in A. The results shown are mean ± standard
deviation from duplicate wells in three independent experiments
(n = 6). Notice the highly significant increase in the
percent survival of EC transfected with A1 or c-IAP2 and exposed to CO
versus EC not exposed to CO (p < 0.0001, unpaired t test).
|
|
Given that HO-1-derived CO activates the p38 MAPK signal transduction
pathway (31) and that the anti-apoptotic effect of HO-1-derived CO is
dependent on the activation of p38 MAPK (31), we tested whether the
"functional interaction" between HO-1/CO and A1 or c-IAP2 (Fig.
8A) required the activation of p38 MAPK. The ability of A1
and c-IAP2 to restore the anti-apoptotic effect of HO-1 was abrogated
when p38 MAPK activation was inhibited by co-expression of a p38
dominant negative mutant (Fig. 9).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 9.
Functional interaction between
NF-B-dependent genes and HO-1
requires the activation of p38 MAPK. 2F-2B EC were transiently
transfected with a -galactosidase reporter (300 ng/3 × 105 cells) and when indicated (+) with IB (500 ng/3 × 105 cells), HO-1 (700 ng/3 × 105 cells) and/or A1 (250 ng/3 × 105
cells) (A), or c-IAP2 (50 ng/3 × 105
cells) (B) expression vectors. When indicated EC were
co-transfected with a p38/CSBP1 dominant negative mutant (p38
MAPK DNM; 250, 500, or 1000 ng/3 × 105 cells)
expression vector. A1 and c-IAP2 expression vectors were used at
suboptimal amounts, which per se do not prevent EC apoptosis
(see Fig. 7). Gray and black histograms represent
EC exposed to medium or TNF- (24 h after transfection; 50 units/ml,
16 h), respectively. Percent survival of transfected EC was
calculated as described under "Experimental Procedures." The
results shown are the mean ± standard deviation from duplicate
wells in three independent experiments (n = 6).
A, notice that the increased percent of survival in EC
expressing HO-1 and A1 is highly significant versus EC
expressing HO-1 without A1 (p < 0.0001, unpaired
t test). The decreased percent of survival in EC expressing
the dominant negative mutant (DNM) of p38 MAPK is highly
significant (p < 0.0001, unpaired t test)
versus EC expressing HO-1 and A1 when the p38 MAPK dominant
negative mutant was expressed at 500-1000 ng/3 × 105
cells. This decrease was not significant (p = 0.2531, unpaired t test) when the p38 MAPK dominant negative mutant
was expressed at 250 ng/3 × 105 cells. B,
increased percent of survival in EC expressing HO-1 and c-IAP2 is
highly significant versus EC expressing HO-1 without c-IAP2
(p < 0.0001, unpaired t test). The decrease
in percent survival in EC expressing the dominant negative mutant
(DNM) of p38 MAPK is highly significant (p < 0.0001, unpaired t test) versus EC expressing
c-IAP2, when the p38 MAPK dominant negative mutant was expressed at 500 ng/3 × 105 cells This decrease was very significant
(p = 0.022, unpaired t test) when the p38
MAPK dominant negative mutant was expressed at 1000 ng/3 × 105 cells but was not significant at 250 ng/3 × 105 cells (p = 0.2054, unpaired
t test).
|
|
| |
DISCUSSION |
In EC the expression of HO-1 can be induced by a multitude of
pro-oxidant stimuli including free heme released from hemoproteins such
as hemoglobin and/or myoglobin when these are oxidized during tissue
injury and necrosis. Free heme intercalates into EC membranes to reach
the intracellular compartment and exert potent cytotoxic effects (29,
44). The only known mechanism by which EC can clear high levels of
intracellular heme is through the up-regulation of HO-1 expression.
Under these circumstances HO-1 becomes the rate-limiting enzyme in the
catabolism of heme into free iron, bilirubin, and CO (25, 26). We have
previously shown that expression of HO-1 is part of a physiological
response to injury by which EC are protected from undergoing apoptosis
(28, 31). We have also shown, along with others, that the
anti-apoptotic effect of HO-1 acts via the generation of CO (31, 45)
and that in EC this anti-apoptotic action depends on the activation of
the p38 MAPK a signal transduction pathway (31).
In the present manuscript we demonstrate that in addition to the
p38 MAPK signal transduction pathway, the anti-apoptotic action of
HO-1 is also dependent on the activation of the transcription factor
NF-B. Once NF-B activation is inhibited, as by
overexpression of its natural inhibitor IB, HO-1 and/or CO
can no longer protect EC cells from undergoing TNF--mediated
apoptosis (Fig. 1). The need for NF-B activation seems specific
to the TNF- signal transduction pathway because CO can protect EC
from serum deprivation-induced apoptosis even if NF-B activation is
inhibited (Fig. 1). Given that we cannot detect significant levels of
NF-B activation by HO-1/CO in EC (Fig. 2), we reasoned that
quiescent EC must have basal levels of NF-B activity that are
required to sustain the anti-apoptotic action of CO. In fact we found
that quiescent EC have significant levels of nuclear NF-
TOP
ABSTRACT
INTRODUCTION
