BIOCHE MICAL AND BIOPHYSICAL RE SE ARCH COMMUNICATIONS 237, 372±381 (1997)
ARTICLE NO. RC977085
17b-Estradiol Inhibits Apoptosis of Endothelial Cells
Rene J . Alvarez J r.,* Sanford J . Gips,* Nicanor Moldovan,* Calvin C. Wilhide,* Emily E. Milliken,*
Arthur T. Hoang,² Ralph H. Hruban,³ Howard S. Silverman,*
Chi V. Dang,² and Pascal J . Goldschmidt-Clermont§
*Division of Cardiology and ²Division of Hematology, Department of Medicine, and ³Departments of Pathology, The J ohns
Hopkins University, Baltimore, Maryland 21287; and §Department of Medicine and Medical Biochemistry,
The Ohio State University, Columbus, Ohio 43210
Received J une 27, 1997
normal vessels, the turnover rate of endothelial cells
is slow (3), and injuries that accelerate endothelial cell
turnover have been shown to alter the ability of these
cells to provide their barrier and metabolic functions
(4). Thus far, there is little information concerning the
control of endothelial cell turnover in vivo.
Apoptosis is a morphologically distinct mechanism
of programmed cell death which involves theactivation
of a cell-intrinsic suicideprogram, and is known toplay
a major role during development, in homeostasis, and
in disease processes (5,6). All mammalian cells possess
the basic machinery to carry out apoptosis, and the
decision to undergo apoptosis is made by individual
cells in responsetointracellular andextracellular stim-
uli. During apoptosis, the nucleus and the cytoplasm
condense toproduce membrane-bound apoptotic bodies
that are phagocytosed by macrophages or adjacent
cells. Apoptosis of endothelial cells is unique in several
ways, including that: (i) endothelial cells are exqui-
sitely sensitive totheir interaction with the extracellu-
lar matrix, whereby disruption of this interaction trig-
gers apoptosis (7,8); (ii) apoptotic arterial endothelial
cells are unlikely to be phagocytosed by adjacent cells;
instead, because of the shear caused by the blood ¯ow,
cells are more likely to detach from the vessel wall
and embolize downstream where their debris can
be processed. With these distinguishing features,
apoptosis of endothelial cells is likely to represent a
major physiological mechanism of endothelial cell
death. Limitedapoptosis of endothelial cells is probably
helping with the maintenance of a functional endothe-
lium by eliminating cells that have become dysfunc-
tional. However, a marked acceleration of theapoptotic
process might lead todisruption of thenormal endothe-
Most women remain free of coronary artery disease
until menopause (9,10). Once the concentration of sex
hormones drops in the plasma of post-menopausal
women, the rate of atherogenesis accelerates progres-
E ndothelial cells provide an antithrombotic and anti-
in¯ammatory barrier for the normal vessel wall. Dys-
function of endothelial cells has been shown to promote
atherosclerosis, and normalization of previously dys-
functional endothelial cells can inhibit the genesis of
atheroma. In normal arteries, endothelial cells are re-
markably quiescent. Acceleration of the turnover rate
of endothelial cells can lead to their dysfunction.
Apoptosis is a physiological process that contributes to
vessel homeostasis, by eliminating damaged cells from
the vessel wall. However, increased endothelial cell
turnover mediated through accelerated apoptosis may
alter the function of the endothelium and therefore, pro-
mote atherosclerosis. Apoptotic endothelial cells can be
detected on the luminal surface of atherosclerotic coro-
nary vessels, but not in normal vessels. This ®nding
links endothelial cell apoptosis and the process of ath-
erosclerosis, although a causative role for apoptosis in
this process remains hypothetical.E strogen metabolites
have been shown to be among the most potent anti-ath-
erogenic agents available to date for post-menopausal
women. The mechanism of estrogen's protective effect
is currently incompletely characterized. Here we show
that 17b-estradiol, a key estrogen metabolite, inhibits
apoptosis in cultured endothelial cells. Our data sup-
port the hypothesis that 17b-estradiol's anti-apoptotic
effect may be mediated via improved endothelial cell
interaction with the substratum, increased tyrosine
phosphorylation of pp125 focal adhesion kinase, and a
subsequent reduction in programmed cell death of en-
dothelial cells. Inhibition of apoptosis by estrogens may
account for some of the anti-atherogenic properties of
? 1997 Academic Press
Endothelial cells have been shown to affect the ho-
meostasis of the vessel wall in terms of vasomotor tone,
platelet andmonocyteadhesion, growth of smooth mus-
cle cells, and extracellular matrix production (1,2). In
Copyright ? 1997 by Academic Press
All rights of reproduction in any form reserved.
Vol. 237, No. 2, 1997BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
sively. Several studies have shown that post-meno-
pausal administration of estrogens inhibits the devel-
opment of coronary artery disease (11,12). Estrogens
may indeed represent the most potent anti-atherogenic
compounds presently available for women (13). How-
ever, long term administration of estrogens post-meno-
pause might be complicated by an increased risk of
breast cancer (14). While studies have suggested that
a target for the protective effect of estrogens is the
endothelium (15), the mechanism by which pre- and
post-menopausal estrogens stabilize endothelial func-
tions remains unknown. We tested the hypothesis that
estrogen stabilization of a functional endothelium in-
volves a mechanism by
apoptosis of endothelial cells in situations of stress. We
also sought the mechanism by which estrogens raise
the threshold for programmed death in these cells.
which estrogens inhibit
APOPTOTIC ENDOTHELIAL CELLS IN DISEASED
Apoptotic endothelial cells were virtually undetect-
able in the coronary vessels of individuals who died of
F IG. 1.
cells appeared retracted and clusters of these cells stained positive for DNA fragmentation by TUNEL. (b) In contrast, apoptotic cells were
virtually undetectable in normal coronary vessel. Endothelial cells forming a con¯uent monolayer were typically ¯at and elongated. Back-
ground staining was with hematoxylin and eosin (magni® ed 1 350). These micrographs are representative of the ® ndings observed in 2
controls and 3 patients with coronary artery disease. These results con® rm a previous report on the association of apoptosis and atheroscle-
rotic plaques (45). (c) Ultrastructural evidence of focal apoptosis within the endothelium of the coronaries of human transplanted heart
with accelerated graft arteriosclerosis. Side by side can be observed (1) a normal appearing endothelial cell; (2) a retracted, but not apoptotic
(perhaps dysfunctional) endothelial cell; (3) a cell displaying several apoptotic characteristics: convolution of the cell surface, vacuole-
formation and budding of the cell surface; (4) a missing cell, possibly apoptotic and dislodged by the shear. This ultrastructural analysis
is representative of the ® ndings observed in 6 patients with accelerated graft arteriosclerosis.
Increased apoptosis of endothelial cells in diseased coronary arteries. (a) At the level of an atherosclerotic plaque, endothelial
F IG. 2.
assay and quantitated by ¯ow cytometry. Bovine aortic endothelial cells were deprived of any estrogens (see Methodology) for 48 hours or
cultured in the same medium, but supplemented with E2. At the end of this incubation, cells were detached with trypsin/EDTA and analyzed
by ¯ow cytometry. A typical histogram from one of three separate experiments is shown. Peak 1 (. . . . ., 104events counted) corresponds to
cells supplemented with E2 (1008M), peak 2 (---, 104events) corresponds to estrogen-starved cells, peak 3 (Ð , 4,991 events) corresponds
to cells maintained in DMEM plus 10% FCS and peak 4 (rrrrr, 104events), same as peak 2, but the terminal deoxyribonucleotidyl
transferase was omitted. (b) To further quantitate the inhibitory effect of E2 on endothelial cell apoptosis in three separate experiments,
the data corresponding tocells exposed toE2 were normalized for the intensity of TUNEL ¯uorescence observed in the absence of estrogens.
The vertical T-bar corresponds to SEM. (c,d) The inhibitory effect of E2 on endothelial cell apoptosis strongly affects the fate of cultured
endothelial cells. Human aortic endothelial cells (from a female donor, passage 7) were estrogen starved (c) or exposed to E2 (d) for 7 days,
then placed back in their regular medium (EGM) for a week. Photographs of representative ® elds are shown.
E2 inhibits apoptosis of cultured endothelial cells: TUNEL assay. (a) Apoptotic cells were stained with the ¯uorescein-TUNEL
Vol. 237, No. 2, 1997BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
F IG. 3.
The DNA of these cells was stained with propidium iodine (PI, red) and phosphatidylserine in the outer lea¯et of the plasma membrane
was detected by binding to FITC-labeled annexin V (green). Overlapping staining of PI and FITC-annexin V appears yellow. While some
cells exposed to E2 did stain positive (a), budding of the plasma membrane, and formation of apoptotic bodies remained limited. In
contrast, endothelial cells deprived of estrogens displayed strong morphological changes consistent with apoptosis (b to d). The vertical line
is 50 mm.
Confocal microscopy. Endothelial cells were either supplemented with E2 [(a), 1008molar], or deprived of estrogens (b to d).
non-atherosclerotic disease (Figure 1a; nÅ3, multiple
vessel slides were studied per individual). In contrast,
endothelial cells examined at the level of atheroscle-
rotic plaques were often retracted and many of these
retracted cells stained positivefor apoptosis (Figure1b,
nÅ3). Ultrastructural analysis was performed on the
coronaries of explanted hearts ® xed immediately after
removal. The hearts were obtained from patients un-
dergoing re-transplantation because of accelerated
graft arteriosclerosis of the ® rst transplant. Scattered
endothelial cells within the coronary vessels displayed
several features that are typical for apoptotic cells, in-
cluding convolution, vacuole-formation and budding of
the cell surface (16) (Figure 1c, nÅ6). In some cells,
we also observed telolysosomes (16) (data not shown).
While apoptotic endothelial cells were seen clustered
at the level of atherosclerotic plaques in classic athero-
sclerosis, in transplant arteriosclerosis, the apoptotic
endothelial cells werefound scattered along thegrafted
endothelium (data not shown).
Since only quiescent endothelial cells are capable of
maintaining the homeostasis of coronary vessels (4),
we theorized that an endothelium containing multiple
apoptotic cells would become dysfunctional and fail to
Vol. 237, No. 2, 1997BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
protect the vessel wall from injuries. Therefore, inhibi-
tors of endothelial cell apoptosis may play a protective
role for the vessel wall and prevent atherosclerotic
to d). We quantitated these apoptotic bodies by ¯ow
cytometric assay, and showed that they were more
abundant in estrogen-deprived endothelial cells (Fig-
ure 4), while ploidy did not seem to be affected by E2,
as nuclei were predominantly 2N with a small fraction
of 4N and virtually no § 8N cells in both estrogen
starved and E2 supplemented conditions.
We also analyzed DNA fragmentation on agarose
gels. Although E2 always inhibited DNA fragmenta-
tion, the structure of the fragments varied somewhat
with the experimental conditions. Early on after estro-
gen withdrawal (within 96 hours), both high molecular
weight (ú30kb) and small internucleosomal DNA frag-
ments were present, while at later time points, only
large fragments were detectable. It is possible that at
later timepoints, remaining cells havea higher thresh-
old for programmed death, and therefore less DNA
fragmentation becomes detectable. Independently of
the time course, the inhibition of DNA fragmentation
by E2 was concentration-dependent (data not shown).
Taken together, these data strongly suggest that the
presenceof theestrogen metaboliteE2, at physiological
concentration, slowed the rate of programmed cell
death and improved the survival of endothelial cells.
17b-ESTRADIOL (E2) AND APOPTOSIS OF
CULTURED ENDOTHELIAL CELLS
To test the effect of a major estrogen metabolite, E2,
on endothelial cell apoptosis, we switched endothelial
cells from their growth medium, which contained estro-
gens, to a medium lacking estrogen metabolites
(õ10012molar, see Methodology). In this estrogen-free
medium, cells were either supplemented with E2 at
indicated concentrations or the vehicle only. The frac-
tion of apoptoticendothelial cells was measuredat time
points varying between 24 and 132 hours of culture in
these conditions, using a battery of assays.
Wefound that exposureof endothelial cells toa phys-
iological concentration of E2 (1008molar) reduced the
number of apoptotic cells. Using terminal dUTP nick
endlabeling (TUNEL) assay toquantitateprogrammed
cell death, the fraction of apoptotic cells was decreased
by 60% (Figure 2a and b). The rate of apoptosis in
estrogen-free medium was greater for senescent endo-
thelial cells (passaged 5 to 10 times) than for younger
cells (passaged 0 to5 times) (17), but therelativeinhib-
itory effect of estrogen supplementation on pro-
grammed cell death was not affected by the number of
passages. The slower apoptotic rate of endothelial cells
provided with E2 does not seem topromotethesurvival
of sick cells, as these cells demonstrate healthy growth
properties. Endothelial cells deprived of estrogens for
seven days then re-exposed to regular medium failed
to recover and continued to deteriorate (Figure 2c). In
contrast, cells that weremaintained in medium supple-
mented with E2 (1008molar) were able toreach con¯u-
ency when switched back to regular medium.
An established step in the process of apoptosis is the
loss of non-equilibrium distribution of phospholipids in
theplasma membrane. Phosphatidylserine, an annexin
V-binding phospholipid, is normally absent from the
outer lea¯et of the plasma membrane. However, in
apoptoticcells, phosphatidylserineis readily detectable
on the surface of cells. To con® rm that endothelial cell
death in estrogen deprived conditions was indeed
mediated by apoptosis, we also stained the cells with
¯uorescein isothiocyanate-labeled annexin-V (FITC-
annexin-V) to detect phosphatidylserine on the outer
lea¯et of the plasma membrane, and with propidium
iodide (PI) to detect chromatin condensation. In the
presence of E2 (1008molar), few endothelial cells were
apoptotic, as shown by annexin-V binding and PI stain-
ing (Figure3a). In contrast, in theabsenceof estrogens,
many morecells displayed signs of apoptosis, including
blebbing of the plasma membrane and release of apop-
totic bodies containing PI-positive material (Figure 3b
EFFECT OF E2 ON ENDOTHELIAL CELL
Endothelial cell survival seems to be exquisitely de-
pendent upon interaction of these cells with proteins
of the extracellular matrix (7,8). Inhibition of endothe-
lial cell binding totheextracellular matrix induces pro-
grammed cell death, a phenomenon which has been
termed anoikis (8). While searching for the mechanism
by which estrogens might reduce programmed death
of endothelial cells, we observed that the presence of
estrogen metabolites potentiated the interaction of en-
dothelial cells with the substratum.
When trypsinized endothelial cells were plated on
cover slips, they spread spontaneously on theglass sur-
face and progressively developed typical focal adhe-
sions, a process which required several hours to com-
plete. We studied the effect of E2 on endothelial cell
spreading (Figure 5). Within 20 minutes after plating
cells on the cover slip, a majority of the cells (60 to
70%) had changed their shape from a rounded con® gu-
ration at the end of the exposure to trypsin, to a ¯at
and spread con® guration. For cells exposed to E2, the
fraction of rounded cells did not increase further, and
typical actin stress ® bers, an index of mature focal ad-
hesions, were readily detectable 30 to60 minutes after
plating the cells. In contrast, in the absence of any
estrogen metabolites, the fraction of rounded cells in-
creased between 30 and 60 minutes, and actin stress
® bers in these rounded cells were markedly reduced
(Figure5). Under our experimental conditions, thefrac-
tion of rounded cells on cover slips deprived of estrogen
Vol. 237, No. 2, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
F IG. 4.
week) and cells supplemented with E2 (1008molar). While cells exposed to E2 (left panel) had trace amount of nuclear microparticles
(marked as M1 bar), estrogen-starved cells contained larger amounts of broken down nuclei (right panel).
Quantitation of apoptotic bodies by ¯ow cytometry. Apoptotic bodies were quantitated in estrogen-starved endothelial cells (one
metabolites was nearly twice that found in cells ex-
posed to estrogens.
One interpretation of our data is that estrogens may
facilitate interaction of endothelial cells with the sub-
stratum, promote the stabilization of focal adhesions
and thereby spreading, and that such effects on cell
adhesion may contribute to reducing anoikis. Alterna-
tively, the effect of E2 might be to inhibit apoptosis
(apoptosis which can induce rounding up of cells)
through a process independent from adhesion, and
therefore enhance cell spreading. Since estrogen depri-
vation does not inducedetectableapoptotic changes be-
fore 24 hours, whereas differences in cell spreading
between E2 supplemented and deprived cells are
readily detectablewithin minutes, wefavor thehypoth-
esis that the improved adhesion observed in the pres-
ence of E2 is causative in the process of inhibition of
strates was not affected by the addition of E2 toplated
endothelial cells, including phospholipase Cg1 (data
not shown). However, we observed that the pp125FAK
of endothelial cells exposed to E2 became increasingly
immunoprecipitable with anti-phosphotyrosine aga-
rose beads, relative to cells deprived of estrogens (Fig-
ure 6a). The band corresponding to pp125FAKwas de-
tected on Western blots with a monospeci® c antibody.
The amount of immunoprecipitable pp125FAKpeaked
at physiological concentrations of E2, and within 60
minutes of exposure to the hormone (data no shown).
The total concentration of this enzyme did not change
after addition of E2 (Figure 6b). To con® rm that tyro-
sine phosphorylation of pp125FAKin cells exposed toE2
was mediated, at least in part, by a mechanism that
involved a functional estrogen receptor, we pre-treated
the endothelial cells with ICI 164-384, a compound
that is a known inhibitor of the estrogen receptor. We
observed that tyrosine phosphorylation of pp125FAKin
response to E2 was decreased (although not com-
pletely) in cells pre-treated with this inhibitor (Fig-
E2 ENHANCES THE PHOSPHORYLATION OF pp125
FOCAL ADHESION KINASE ON TYROSINE
Thepp125 focal adhesion kinase(pp125FAK) is associ-
ated with focal adhesions and becomes phosphorylated
on tyrosine upon integrin engagement. If estrogens im-
prove endothelial cell adhesion, the tyrosine phosphor-
ylation of pp125FAKshould re¯ect this effect. To test
this hypothesis, endothelial cells were deprived of es-
trogens in culture for 48 hours, and then exposed to
serial concentrations of E2, as indicated (Figure 6). We
measured the tyrosine phosphorylation of pp125 focal
adhesion kinase (pp125FAK) by immunoprecipitation
and Western blotting assay.
The phosphorylation of most tyrosine kinase sub-
Our data support the hypothesis that a major estro-
gen metabolite, E2, at physiological concentration in-
hibits endothelial cell apoptosis in culture. While the
mechanism responsible for the anti-apoptotic effect of
E2 remains to be established, our data support the
hypothesis that E2 promotes interaction of endothelial
cell integrins with the substratum, thereby enhancing
the tyrosine phosphorylation of the focal adhesion ki-
nase(18) and reducing endothelial cell anoikis (7,8). By
Vol. 237, No. 2, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
F IG. 5.
for 48 hours, then trypsinized and plated on cover slips in the presence (b,e) or the absence (a,d) of E2. Cells were ® xed at various time
points following plating of the cells and stained for ® lamentous actin (a,b; rhodamine phalloidin) or tyrosine phosphorylated substrates
(d,e; ¯uorescein). For each time point, spread cells were counted by an investigator blinded with regard to the cell treatment (c). This
experiment is representative of three separate experiments. The bar is 5 mm.
E2 enhances the spreading of endothelial cells. Human endothelial cells (female donor, passage 5) were starved of estrogens
F IG. 6.
developed for pp125FAK. Bovine aortic endothelial cells were starved of estrogens for 48 hours, then exposed to various concentrations of
E2 for onehour. At theendof this exposure, thecells werelysedand solubilizedtyrosinephosphorylatedsubstrates wereimmunoprecipitated.
The immune complexes were analyzed on Westerns developed with an anti-pp125FAKmonoclonal antibody. The two lanes on the right
correspond to cells that have been pre-treated with ICI 164-384 (2 mg/ml, for 24 hours). (b) Western blot of normalized extracts obtained
from endothelial cells that were either estrogen-starved (time 0) or exposed to E2 (1008molar) for 5, 10, 60 and 120 min, and developed
with a monoclonal antibody against pp125FAK. Note that exposure to E2 did not affect the cellular concentration of pp125FAK.
Tyrosine phosphorylation of pp125FAKis stimulated by E2. (a) Western blot of anti-phosphotyrosine-precipitated proteins,
Vol. 237, No. 2, 1997BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
reducing the rateof endothelial cell apoptosis, estrogen
metabolites may stabilize the endothelium in a quies-
cent state. Quiescent endothelial cells are poised to (i)
titrate the vessel tone appropriately, (ii) prevent plate-
let adhesion and aggregation, as well as adhesion of
circulating leukocytes and other immune competent
cells, (iii) down regulate the coagulation cascade and
(iv) titrate the proliferation of smooth muscle cells and
other cells of the vessel wall to exactly compensate for
the death of these cells occurring at steady state.
Since the presentation of our preliminary data on
E2 and endothelial cell apoptosis (19), a report from
another group has con® rmed our original observation
(20). Moreover, the enhancing effect of estrogens on
endothelial cell adhesion has been linked to the ability
of estrogens to promote the angiogenic potential of en-
dothelial cells (15). The protective effect of estrogens
towards endothelial cell apoptosis may also provide a
mechanism for theangiogenic effect of estrogens. Thus,
upon improving adhesion and reducing programmed
death of endothelial cells, estrogens might enhance the
probability that migrating endothelial cells will suc-
cessfully colonize ischemic tissues. In contrast with the
angiogenic effect of estradiol, another estrogen metabo-
lite, 2-methoxyestradiol, is able toinhibit the prolifera-
tion of endothelial cells andthereby suppress angiogen-
esis and tumor growth (21). It is possible that 2-meth-
oxyestradiol can counteract theaction of other estrogen
metabolites through an unknown mechanism. To clar-
ify these seemingly opposite effects, further experi-
ments are needed to characterize the mechanism by
which the estrogen receptor regulates focal adhesions
and the reactions, perhaps downstream from pp125FAK
activation, that lead to inhibition of endothelial cell
Estrogen deprivation has been shown to induce
apoptosis of estrogen receptor positive MCF-7 breast
cancer cells (22), and estrogen supplementation inhib-
its apoptosis of ovarian granulosa cells (23). Because
endothelial cells, like ovarian granulosa cells and the
estrogen receptor positive breast cancer cell line MCF-
7 cells, contain relatively large concentrations of estro-
gen receptor molecules (24,25), it is possible that for
such cells both proliferation and programmed death is
tightly regulated by estrogen metabolites.
The action of estrogen metabolites is believed to be
mediated by the genomic effects of the estrogen recep-
tor. In the absence of hormone, the estrogen receptor
is found as an ` ` aporeceptor complex'', consisting of a
receptor monomer, a dimer of the heat shock protein
Hsp90, and additional ligand proteins (26-30). Upon
binding to its estrogen-ligand, the ` ` activated'' receptor
dissociates from the cytoplasmic protein complex, di-
merizes, binds tospeci® cpalindromicDNA targets, and
thereby modulates the transcription of selected genes.
It is believed that it is through changing protein con-
centrations via genomic regulation that estrogens in-
duce their effects in target cells (31). However, consid-
ering therapideffect of E2 on thetyrosinephosphoryla-
tion of pp125FAK, it is also possible that some effects of
estrogen metabolites might not be mediated solely by
the product of protein synthesis. Instead, the cyto-
plasmic interaction of the estrogen receptor with heat
shock proteins and additional ligands, an interaction
that is disrupted by the binding to estrogen metabo-
lites, couldmediatesomeof therapideffects of estrogen
metabolites on endothelial cells (26).
Estrogens have been shown to exert a vasodilatory
effect on the coronary vessels of animals and humans
studied in the catheterization laboratory (32-34). The
vasodilatation induced by estrogen metabolites was
shown to occur within minutes after exposure to the
steroid hormone. Such an effect could result from very
rapid genomic regulation of gene expression by the es-
trogen receptor. However, the vasodilatory effects of
estrogens may be also mediated by estrogen-triggered
post-translational signalling reactions occurring dur-
ing the cytoplasmic life span of the estrogen receptor,
similar to those postulated for the tyrosine phosphory-
lation of pp125FAK. Further studies will be needed to
delineate precisely the role of the estrogen receptor in
theseresponses. Therecent discovery of a secondrecep-
tor for estrogens (ER-b) (35), which might mediate, at
least in part, the effects of estrogen metabolites on the
vessel wall (36), could account for the lack of major
impact of the disruption of the ER-a on mammalian
We conclude that the effect of E2 on endothelial cell
adhesion and apoptosis may have substantial conse-
quences for the vessel wall. The concurrent improve-
ment of circulating lipids (9), the redox milieu (34,38),
and rate of endothelial cell apoptosis by estrogens may
account for the remarkable protection against athero-
sclerosis that women experience prior to menopause.
Histology and transmission electron microscopy.
slides, the coronary vessels were ® xed immediately in fresh 4% (wt/
vol) formaldehyde, transferred to phosphate buffered saline (PBS)
solution and embedded in paraf® n. The 4-mm sections were stained
with hematoxylin and eosin, and for apoptosis using the TUNEL
assay (39). Brie¯y, thecoronary sections wereextractedfromparaf® n
by incubation in xylene (3 times for ® ve minutes), 100% ethanol
(twice for 2 minutes), 95% ethanol (twice for 2 minutes), and in
PBS for 5 minutes. Nuclear proteins were hydrolyzed by exposure
to proteinase K for 10 minutes (20 mg/ml). After two rinses in water
(5 minutes each) and one rinse in 2% hydrogen peroxide (5 minutes,
to quench the endogenous peroxidase), apoptotic cells were stained
by direct immunoperoxidase detection of digoxigenin-labeled geno-
mic DNA, using theApopTag In Situ peroxidasedetection kit (Oncor,
Inc., Gaithersburg, MD) and the manufacturer's instructions. Con-
trols includedcoronary sections processedin a similar fashion, except
that the terminal deoxynucleotidyl transferase enzyme was omitted.
For electron microscopy, the coronaries of patients undergoing or-
thotopic heart transplantation were ® xed in the operation room with
2.5% glutaraldehyde (pH: 7.3). Further processing of the replicates
was performed as reported (40).
Vol. 237, No. 2, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Cell culturemedia and reagents.
from cow aorta by collagenase hydrolysis and grown in plastic ¯asks
in DMEM supplementedwith 10% fetal calf serum(FCS), in a humid-
i® ed atmosphere of 5% CO2/95% air at 37?C. Human aortic endothe-
lial cells werepurchasedfrom Clonetics Corporation (San Diego, CA),
and grown in endothelial cell growth medium with 2% FCS (EGM).
The concentration of total estrogens in DMEM with 10% FCS was
21 ng l01, and in EGM was 13 ng l01. Totest for theeffect of estrogens
on endothelial cell apoptosis and tyrosine phosphorylation, bovine
and human endothelial cells were transferred to a medium of very
low estrogen content (LEM), consisting of DMEM without phenol red,
supplemented with 10% castrated horse serum. The total estrogen
concentration of LEM was õ10012M (õ2 ng l01. After 48 hours
in this medium, endothelial cells were either supplemented with
indicated concentrations of E2 or with the vehicle only (1003% etha-
Endothelial cells wereharvestedsine phosphorylated proteins were immunoprecipitated with beads
coated with anti-phosphotyrosine antibodies. The extracts were nor-
malized for total protein content prior to the addition of anti-phos-
photyrosine agarose beads. Immunoprecipitated proteins were ana-
lyzed on SDS-Page, and individual Western blots were developed
with a monoclonal anti-pp125FAKantibody, or another monoclonal
antibody against phosphotyrosine residues to detect all tyrosine
phosphorylated proteins (the antibodies were from Upstate Biotech-
nology Inc., Lake Placid NY). For total pp125FAKcontent, Western
blots of normalized cell extracts were developed with a monoclonal
antibody against pp125FAK.
This work was supported by the Bernard Foundation and by the
Scleroderma Research Foundation.
Assays for apoptosis of cultured aortic endothelial cells.
¯uent bovine aortic endothelial cells were exposed to LEM, or to
LEM supplemented with E2 (1008M) for 48 to 96 hours. At the end
of this incubation, cells were detached by exposure to trypsin (41),
and apoptotic cells were ¯uorescently labeled using the ¯uorescein
ApopTag apoptosis detection kit from Oncor, and the instructions
provided by the manufacturer. Fluorescein was detected by ¯ow cy-
tometry (FACScan, Becton Dickinson). (ii) Laser scanning confocal
microscopy was used to detect the morphological characteristics of
apoptotic cells. Using a BioRad MRC 600, with a BioRad GHS ® lter
block, sequential images (z-series) corresponding to slices approxi-
mately 1.5 mm thick and 1 mm apart in the z-axis were obtained
(42). The presence of phosphatidylserine in the outer lea¯et of the
plasma membrane was detected with the binding of FITC-annexin
V. DNA material was detected by propidium iodide staining (PI, 10
mg ml01) (43). Images were stored tooptical disc, and false color hard
copies were generated on a color video printer (42). (iii) Ploidy and
nuclear fragmentation was also quantitated by staining human aor-
tic endothelial cells with PI, and measuring ¯uorescence by ¯ow
cytometric assay (43). (iv) DNA fragmentation and the presence of
large or oligonucleosomal DNA fragments of (180)nbase pairs was
assessed using gel electrophoresis of extracted DNA and staining of
the gels with ethidium bromide (44).
1. Vane, J . R., Anggard, E. E., and Botting, R. M. (1990) New Engl.
J . Med. 323, 27±36.
2. Dzau, V. J ., and Gibbons, G. H. (1991) Hypertension 18(Suppl.
3. Hanahan, D., and Folkman, J . (1996) Cell 86, 353±364.
4. Shimokawa, H., Flavahan, N. A., and Vanhoutte, P. M. (1990)
Circ. Res. 65, 740±753.
5. Martin, S. J ., Green, D. R., and Cotter, T. G. (1994) TIBS 19,
6. Steller, H. Mechanisms and genes of cellular suicide. (1995) Sci-
ence 267, 1445±1448.
7. Meredith, J . E., Fazeli, B., and Schwartz, M. A. (1993) Mol. Biol.
Cell 4, 953±961.
8. Re, F., Zanetti, A., Sironi, M., Polentarutti, N., Lanfrancone, L.,
Dejana, E., and Colotta, F. (1994) J . Cell Biol. 127, 537±546.
9. Bush, T. L. (1990) Ann. N.Y. Acad. Sci. 592, 263±271.
10. Matthews, K. A., Meilahn, E., Kuller, L. H., Kelsey, S. F., Caggi-
ula, A. W., and Wing, R. R. (1989) New Engl. J . Med. 321, 641±
11. Stampfer, M. J ., Colditz, G. A., Willett, W. C., Manson, J . E.,
Rosner, B., Speizer, F. E., and Hennekens, C. H. (1991) New
Engl. J . Med. 325, 756±762.
12. Nabulsi, A. A., Folsom, A. R., White, A., Patsch, W., Heiss, G.,
Wu, K. K., and Szklo, M. (1993) New Engl. J . Med. 328, 1069±
13. Barrett-Connor, E., and Bush, T. L. (1991) J AMA 265, 1861±
14. Grodstein, F., Stampfer, M. J ., Colditz, G. A., Willett, W. C.,
Manson, J . E., J offe, M., Rosner. B., Fuchs, C., Hankinson, S. E.,
Hunter, D. J ., Hennekens, C. H., and Speizer, F. E. (1997) New
Engl. J . Med. 336, 1769±1775.
15. Morales, D. E., McGowan, K. A., Grant, D. S., Maheshwari, S.,
Bhartiya, D., M. C., Cid, M. C., Kleinman, H. K., and Schnaper,
H. W. (1995) Circulation 91, 755±763.
16. Kerr, J . F. R. (1995) Trends Cell Biol. 5, 55±57.
17. Varani, J ., Dame, M. K., Taylor, C. G., Sarma, V., Merino, R.,
Kunkel, R. G., Nunez, G., and Dixit, V. M. (1995) Lab. Invest.
18. Ruoslahti, E., and Reed, J . C. (1994) Cell 77, 477±478.
19. Gips, S. J ., Alvarez, R. J ., Milliken, E. E., Dang, C. V., and
Goldschmidt-Clermont, P. J . (1996) J . Am. Coll. Cardiol. 27,
20. Spyridopoulos, I., Sullivan, A. B., Kearney, M., Isner, J . M., and
Losordo, D. W. (1997) Circulation 95, 1505±1514.
Cell spreading assay.
were exposed to LEM for 48 hours. After this incubation, cells were
detached from the ¯asks by exposure to Trypsin-EDTA for two min-
utes, and then resuspended in 2 ml of LEM. The cell suspension was
then split into two aliquots (1 ml, 106cells each). One aliquot was
supplemented with E2 (1008M) while the other was supplemented
with the vehicle alone (1003% ethanol, vol/vol). A droplet (75 ml) of
each aliquot was placed individually in the center of three separate
glass cover slips. The cells were then incubated at 37?C, in 5% CO2
in air for twenty minutes. At the end of this incubation, each cover
slip was rinsed twice with phosphate buffered saline. Then, 2 ml of
LEM supplemented with either E2 or vehicle, were added to the
estrogen supplemented andestrogen starved cells, respectively. Cells
on cover slips were ® xed in 3.7% formaldehyde after up to 80 min
incubations at 37?C, 5% CO2. The cells were permeabilized with
Triton X 100 (0.2%), and stained for ® lamentous actin (with rhoda-
mine phalloidin), and for tyrosine phosphorylated proteins (with
anti-phosphotyrosine as primary antibody and ¯uorescein isothiocy-
anate labeled secondary antibody) as reported (41). The coverslips
were mounted, analyzed (Zeiss microscope), and photographed (Ko-
dak Ektachrome 400X ® lms) (41).
Con¯uent human aortic endothelial cells
Tyrosine phosphorylation of pp125FAK.
cells were placed in LEM for 48 hours prior to exposure to various
concentrations of E2. At the end of 60 min, the cells were extracted
in lysis buffer (15 mM Hepes, pH 7.0, 145 mM NaCl, 0.1 mM MgCl2,
1 mM Na vanadate, 10 mM EGTA, 0.5% Triton X 100, 1 mM 4-2-
aminoethyl-benzenesulfonyl¯uoride, 20 mg/ml of proteaseinhibitors:
chymostatin, leupeptin, antipain, pepstatin) (41). Solubilized tyro-
Bovine aortic endothelial
Vol. 237, No. 2, 1997 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Download full-text
21. Fotsis, T., Zhang, Y., Pepper, M. S., Adlercreutz, H., Montesano,
R., Nawroth, P. P., and Schweigerer, L. (1994) Nature368, 237±
22. Kyprianou, N., English, H. F., Davidson, N. E., and Isaacs, J . T.
(1991) Canc. Res. 51, 162±166.
23. Billig, H., Furuta, I., and Hsueh, A. J . W. (1993) Endocrinology
24. Post, W. S., Goldschmidt-Clermont, P. J ., Heldman, A. W., Suss-
man, M. S., Ouyang, P., Milliken, E. E., and Issa, J .-P. J . (1997)
25. Venkov, C. D., Rankin, A. B., and Vaughan, D. E. (1996) Circula-
tion 94, 727±733.
26. Bohen, S. P., Kralli, A., and Yamamoto, K. R. (1995) Science268,
27. Picard, D., Khursheed, B., Garabedian, M. J ., Fortin, M. G.,
Lindquist, S., and Yamamoto, K. R. (1990) Nature348, 166±168.
28. Bohen, S. P., and Yamamoto, K. R. (1993) Proc. Natl. Acad. Sci.
USA. 90, 11424±11428.
29. Tsai, M. J ., and O'Malley, B. W. (1994) Annu. Rev. Biochem. 63,
30. Lubahn, D. B., Moyer, J . S., Golding, T. S., Couse, J . F., Korach,
K. S., and Smithies, O. (1993) Proc. Natl. Acad. Sci. USA. 90,
31. Evans, R. M. (1988) Science 240, 889±895.
32. Williams, J . K., Adams, M. R., and Klopfanstein, H. S. (1990)
Circulation 81, 1680±1687.
33. Reis, S. E., Gloth, S. T., Blumenthal, R. S., Resar, J . R., Zacur,
H. A., Gerstenblith, G., and Brinker, J . A. (1994) Circulation 89,
34. Keaney, J . F. J r., Shwaery, G. T., Xu, A., Nicolosi, R. J ., Loscalzo,
J ., Foxall, T. L., and Vita, J . A. (1994) Circulation 89, 2251±
35. Kuiper, G. G., Enmark, E., Pelto-Huikko, M., Nilsson, S., and
Gustafsson, J . A. (1996) Proc. Natl. Acad. Sci. USA. 93, 5925±
36. Iafrati, M. D., Karas, R. H., Aronovitz, M., Kim, S., Sullivan, J r.,
T. R., Lubahn, D. B., O'Donnell, J r., T. F., Korach, K. S., and
Mendelsohn, M. E. (1997) Nature Med. 3, 545±548.
37. Korach, K. S. (1994) Science 266, 1524±1527.
38. Subbiah, M. T., Kessel, B., Agrawal, M., Rajan, R., Abplanalp,
W., and Rymaszewski, Z. (1993) J . Clin. Endoc. Metab. 77, 1095±
39. Gavrieli, Y., Sherman, Y., and Ben-Sasson, S. A. (1992) J . Cell.
Biol. 119, 493±501.
40. Young-Bamsaran, J . O., Hruban, R. H., Hutchins, G. M., Phelps,
T. H., Baumgartner, W. A., Reitz, B. A., and Olson, J . L. (1993)
Ultrastruc. Pathol. 17, 125±136.
41. Finkel, T., Theriot, J . A., Dise, K. R., Tomaselli, G. F., and
Goldschmidt-Clermont, P. J . (1994) Proc. Natl. Acad. Sci. USA.
42. Crawford, L. E., Tucker, R. W., Heldman, A. W., and Gold-
schmidt-Clermont, P. J . (1994) Adv. Exp. Med. Biol. 358, 105±
43. Greenberg, S. M., Rosenthal, D. S., Greeley, T. A., Tantravahi,
R., and Handin, R. I. (1988) Blood 72, 1968±1977.
44. Hoang, A. T., Cohen, K. J ., Barrett, J . F., Bergstrom, D. A., and
Dang, C. V. (1994) Proc. Natl. Acad. Sci. USA. 91, 6875±6879.
45. Isner, J . M., Kearney, M., Bortman, S., and Passeri, J . (1995)
Circulation 91, 2703±2711.