Estrogen and Tamoxifen Metabolites Protect Smooth Muscle
Cell Membrane Phospholipids Against Peroxidation and
Inhibit Cell Growth
Raghvendra K. Dubey, Yulia Y. Tyurina, Vladimir A. Tyurin, Delbert G. Gillespie, Robert A. Branch,
Edwin K. Jackson, Valerian E. Kagan
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Abstract—The goal of this study was to test the hypothesis that antioxidant estrogens, by a mechanism independent of the
estrogen receptor, protect phospholipids residing in the plasma membrane of vascular smooth muscle cells from
peroxidation and peroxidation-induced cell growth and migration. Peroxidation of membrane phospholipids was
assessed by HPLC analysis of phospholipids extracted from rat aortic vascular smooth muscle cells prelabeled with
cis-parinaric acid (a fatty acid that is susceptible to peroxidation, which quenches its fluorescent properties). Incubation
of cells for 2 hours with the peroxyl radical donor 2,29-azobis-2,4-dimethylvaleronitrile (AMVN) caused peroxidation
of all measured membrane phospholipids. This effect was attenuated by pretreating cells for 15 minutes with 50 to 5000
ng/mL of 2-hydroxyestradiol (strong antioxidant but weak estrogen-receptor ligand) or 4-hydroxytamoxifen (strong
antioxidant and potent estrogen-receptor ligand), but not by estrone or droloxifene (both weak antioxidants but potent
estrogen-receptor ligands). Moreover, pretreatment of cells for 20 hours with physiological concentrations (0.3 ng/mL)
of 2-hydroxyestradiol or pharmacologically relevant concentrations of 4-hydroxytamoxifen (40 ng/mL) also decreased
AMVN-induced phospholipid peroxidation. Both 2-hydroxyestradiol and 4-hydroxytamoxifen were as effective as
2,2,5,7,8-pentamethyl-6-hydrochromane (an antioxidant homolog of vitamin E) in attenuating AMVN-induced
peroxidation of membrane phospholipids. Also, physiological concentrations of 2-hydroxyestradiol, but not estrone, and
pharmacologically relevant concentrations of 4-hydroxytamoxifen attenuated AMVM-induced DNA synthesis, cell
proliferation, and cell migration. These studies demonstrate in vascular smooth muscle cells that antioxidant estrogens
via a non-estrogen receptor– dependent mechanism attenuate peroxidation of membrane phospholipids and
peroxidation-induced cell growth and migration. (Circ Res. 1999;84:229-239.)
Key Words: estrogen n metabolism n antioxidant n cardiovascular disease n free radical
A
lthough it is well established that estrogens are cardioprotective in postmenopausal women,1–5 the mechanisms of estrogen-induced reduction in the risk of coronary
artery disease remain unclear. Because a complete understanding of estrogen-mediated cardioprotection may allow
the development of novel agents for the prevention and
treatment of coronary artery disease in both women and men,
further exploration in this regard is warranted.
One explanation that has been advanced to account in part
for the cardioprotective effects of estrogens is the antioxidant
hypothesis. Two main lines of evidence support this hypothesis. First, estrogens or their metabolites contain 1 or more
phenolic functional groups, and the phenol moiety can
scavenge free radicals.6 For example, 17b-estradiol not only
contains a phenolic functional group, but also is metabolized
to 2-hydroxyestradiol, a catechol estrogen and potent antiox-
idant. Although tamoxifen, another cardioprotective
estrogen-receptor ligand,7 does not contain a phenolic group,
it can be metabolized to the antioxidant phenolic compound
4-hydroxytamoxifen.8 Second, estrogen and tamoxifen attenuate free radical–induced oxidation of plasma LDL,8 –13
although diminished formation of oxidized LDL per se
cannot explain all the cardiovascular benefits of estrogens,
because restenosis after angioplasty and bypass surgery, a
process that is independent of oxidized LDL, is attenuated by
estrogens.14 –17
It is conceivable that estrogens could also protect critical
cellular structures from oxidative damage and that this may
contribute to the cardioprotective effects of estrogens. In this
regard, it is well established that free radicals oxidize membrane lipids and integral membrane proteins18,19 and that such
oxidations can alter signal-transduction mechanisms control-
Received July 28, 1998; accepted November 11, 1998.
From the Center for Clinical Pharmacology and Department of Medicine (R.K.D., D.G., R.A.B., E.K.J.), Department of Pharmacology (E.K.J.),
Department of Environmental and Occupational Health (Y.Y.T., V.A.T., V.E.K.), and Cancer Institute (V.E.K.), University of Pittsburgh, Pittsburgh, Pa;
and Clinic for Endocrinology, Department of Obstetrics and Gynecology (R.K.D.), University Hospital Zurich, Switzerland.
Correspondence to Dr Raghvendra K. Dubey, Center for Clinical Pharmacology, Department of Medicine and Pharmacology, 623 Scaife Hall, 200
Lothrop St, University of Pittsburgh Medical Center, Pittsburgh, PA 15213-2582. E-mail [email protected]
© 1999 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
229
230
Estrogen Prevents Membrane Peroxidation and Growth
ling cell migration and proliferation,20,21 processes critical to
vascular neointimal formation. Moreover, because estrogens
are lipophilic, high local concentrations of these compounds
would exist in the biophase of the lipid bilayer during
hormone replacement therapy. Thus, the hypothesis that
estrogens may attenuate oxidation of membrane components
has a strong a priori rationale. Accordingly, the main objective of the present study was to test the hypothesis that
antioxidant estrogens, by a mechanism independent of the
estrogen receptor, protect phospholipids residing in the
plasma membrane of vascular smooth muscle cells from free
radical–induced peroxidation and peroxidation-induced cell
growth and migration.
Materials and Methods
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
cis-Parinaric acid (Z-9,E-11,E-13,Z-15– octadecatetraenoic acid) was
purchased from Molecular Probes, Inc. The purity of each batch of
cis-parinaric acid used was determined by UV spectrophotometry
(Shimadzu UV 160U spectrophotometer) using a molar extinction at
304 nm in ethanol of 803103 (mol/L)21 z cm21. 2,29-Azobis (2,4dimethylvaleronitrile) (AMVN) was procured from Polysciences,
Inc. 2,29-Azobis (2-aminodinopropane) dihydrochloride (AAPH)
was purchased from Wako Chemicals USA Inc. Fatty acid-free
human serum albumin (hSA), phospholipids, 5,59-dithiobis(2nitrobenzoic acid), estrone, glutathione, glutathione-peroxidase,
cumene hydroperoxide, 2-mercaptoethylamine (cysteamine) hydrochloride, Tween 20, sodium molybdate, malachite green base,
butylated hydroxytoluene, and 3-[4,5-dimethylthiozol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) were purchased from Sigma
Chemical Co. Methanol, chloroform, hexane, 2-propanol, and water
were purchased from Aldrich Chemical Co. Deferoxamine mesylate
was purchased from Ciba. 2,2,5,7,8-Pentamethyl-6-hydrochromane
(PMC) was a generous gift from Eisai Co. 4-Hydroxytamoxifen and
droloxifene were obtained from Research Biochemicals International. 2-Hydroxyestradiol was purchased from Steraloids Inc. Phenol red–free DMEM and DMEM/F12 medium, HBSS, penicillin,
streptomycin, 0.25% trypsin-EDTA solution, and all tissue culture
instruments were purchased from Gibco Laboratories. FCS was
obtained from HyClone Laboratories Inc. All other chemicals used
were of tissue-culture grade or best grade available.
General Experimental Approach
To test the hypothesis that antioxidant estrogens inhibit oxidation of
membrane structures, it was necessary to measure free radical–
induced oxidation of membrane lipids in intact vascular smooth
muscle cells. In this regard, we used the novel strategy of monitoring
fluorescence intensity of membrane-incorporated cis-parinaric acid,
a naturally occurring 18-carbon fatty acid that is highly susceptible
to free radical–induced peroxidation, which markedly quenches its
fluorescence.22 Because cis-parinaric acid fluorescence is lost permanently on oxidation of the conjugated double-bond system, repair
mechanisms, which are normally present in live cells, cannot mask
the oxidative events. This is true even if cis-parinaric acid is oxidized
and replaced with unoxidized fatty acids, since the evidence for
oxidation remains as loss of cis-parinaric acid fluorescence. This
characteristic is not found in assays currently used in cultured cells
and animal models. Furthermore, use of cis-parinaric acid also
enables the localization of the effects of oxidative stress on different
classes of phospholipids. Peroxyl radicals were generated by the
addition of AMVN,23 and fluorescence intensity of the reporter
molecule, cis-parinaric acid, was monitored in several membrane
phospholipid components in the absence and presence of
2-hydroxyestradiol, 4-hydroxytamoxifen, estrone, and droloxifene.
These particular estrogens were selected because 2-hydroxyestradiol
and 4-hydroxytamoxifen are strong antioxidants but different in their
potency as estrogen-receptor ligands (4-hydroxytamoxifen.2hydroxyestradiol), whereas estrone and droloxifene are weak antioxidants but potent estrogen-receptor ligands.6,8,11 We anticipated
that using a diverse set of estrogens with different antioxidant and
estrogen-receptor ligand affinities would allow inferences regarding
whether a given effect was dependent or independent of estrogen
receptors. Moreover, we used PMC, a potent antioxidant, as a
reference benchmark for evaluating the antioxidant potency of the
estrogens tested in this study. We also investigated the selected
estrogens with regard to their ability to attenuate peroxyl radical–
induced changes in endogenous thiols (reduced glutathione [GSH]
and protein sulfhydryls) in vascular smooth muscle cells, using in
this case AAPH23 as the oxidant stress. The addition of these
experiments allowed us to contrast the antioxidant effects of estrogens at 2 important targets for oxidative changes. Finally, to evaluate
the physiological relevance of the antioxidant effects of estrogens,
we conducted cell proliferation, [3H]thymidine incorporation, and
cell migration studies to determine whether peroxyl radicals induce
smooth muscle cell growth and migration and whether the selected
estrogens inhibit free radical–induced DNA synthesis, cell proliferation, and cell migration.
Aortic Smooth Muscle Cell Culture
Male Sprague-Dawley rats (Charles River, Wilmington, MA) weighing 150 to 200 g were fed standard rat chow and tap water ad libitum.
Aortic smooth muscle cells were cultured as explants from the lower
abdominal aortas that were obtained from ether-anesthetized rats
after a midline abdominal incision including the diaphragm and as
described previously.24 Smooth muscle cell purity was characterized
by immunofluorescence staining with specific anti–smooth muscle
a-actin monoclonal antibodies and by morphological criteria specific
for smooth muscle as described in detail previously.25 Smooth
muscle cells between the second and third passages were used for the
cellular lipid peroxidation and thiol (GSH) oxidation studies.
Incorporation of cis-Parinaric Acid Into Smooth
Muscle Cell Phospholipids
cis-Parinaric acid was incorporated into smooth muscle cells by
addition of cis-parinaric acid complexed with hSA [cis-parinaric acid
(500 mg, 1.8 mmol) in 25 mL of DMSO and hSA (50 mg, 760 nmol)
in 1 mL of PBS (in mmol/L, 137 NaCl, 2.7 KCl, 1.5 KH2PO4, and 8
Na2HPO4, pH 7.4)] to the cell suspensions. Briefly, confluent
monolayers of smooth muscle cell were dislodged by trypsinization,
washed once with DMEM containing 10% FCS, and washed
subsequently with DMEM alone. Before the experiments, the smooth
muscle cells were rinsed twice with L1210 buffer containing
(in mmol/L) 115 NaCl, 5 KCl, 1 MgCl2, 5 NaH2PO4, 10 glucose, and
25 HEPES (pH 7.4). Cells were diluted to a density of 1.03106
cells/mL and then incubated with cis-parinaric acid-hSA complex
(final concentration of 5 mg/mL cis-parinaric acid) in L1210 buffer
at 37°C in the dark and under air. The smooth muscle cells were
incubated for different time periods (30, 60, 120, and 180 minutes),
and the time course of cis-parinaric acid incorporation into cellular
phospholipids was evaluated after washing the cells twice with
L1210 buffer with and without hSA (0.5 mg/mL). Trypan blue
exclusion tests for cell viability were conducted in parallel by taking
aliquots of smooth muscle cells treated for different time periods.
AMVN-Induced Peroxidation of Membrane Lipids
AMVN, a lipid-soluble azo- compound that is a thermally activated
initiator of peroxyl radicals, was used to induce lipid peroxidation.
Because of its hydrophobic character, AMVN partitions into the acyl
lipid region of membranes and generates radicals26 that do not escape
from the hydrophobic lipid environment.23 Moreover, oxidation of
phospholipids by AMVN tends to be nonselective, so in principle it
is possible to determine whether particular phospholipid classes are
intrinsically more susceptible to oxidation than others. Smooth
muscle cells were incubated for 2 hours with cis-parinaric acid, and
cis-parinaric acid–labeled cells were subsequently incubated in the
presence of AMVN (500 mmol/L) in L1210 buffer at 37°C in the
dark in air for 2 hours. Appropriate amounts of phenolic compounds
(4-hydroxytamoxifen, droloxifene, estrone, 2-hydroxyestradiol, or
PMC) or cysteamine were added to the incubation medium 15
Dubey et al
February 5, 1999
231
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
minutes before addition of AMVN. After incubation for 2 hours,
aliquots of cell suspension were taken for both determination of cell
viability and lipid analysis.
To evaluate whether long-term treatment of cells with
4-hydroxytamoxifen, estrone, 2-hydroxyestradiol, and PMC prevents
AMVN-induced peroxidation of smooth muscle cells, trypsinized
cells were suspended in DMEM containing 0.4% albumin and
treated for 20 hours with or without 0.3 to 30 ng/mL of the above
agents. Following incubation the smooth muscle cells were loaded
with cis-parinaric acid (as described above) and subsequently incubated with AMVN (250 mmol/L) for 3 hours. After incubation,
aliquots of cell suspension were taken for both determination of cell
viability and lipid analysis.
mixture of chloroform, acetone, methanol, glacial acetic acid, and
water (50:20:10:10:5, vol/vol). The phospholipids were then visualized by exposing the plates to iodine vapor and the spots identified
by comparison with migration of authentic phospholipid standards.
The spots identified by iodine staining were scraped off the plates,
and the silica acid was transferred to tubes. Lipid phosphorus in
individual phospholipid spots of the HPLC plates were extracted and
quantified spectrophotometrically with the method of Arduini et al.30
The identity of each phospholipid was established by comparison
with the Rf values measured for authentic standards.
Extraction of Cell Lipids
AMVN-derived peroxyl radicals do not escape the lipid environment
of the membrane23; hence, AMVN is not a good agent with which to
study the effects of peroxyl radicals on intracellular glutathione
oxidation. Therefore, we used AAPH, an analog of AMVN that is a
water-soluble azo- initiator. Smooth muscle cells were incubated
with AAPH (20 mmol/L in L1210 buffer, pH 7.4) for 2 hours at 37°C
in the absence or in the presence of different concentrations of the
phenolic compounds (4-hydroxytamoxifen, droloxifene, estrone,
2-hydroxyestradiol, or PMC). The reaction was terminated by
cooling the samples to 4°C. In parallel to the above vascular smooth
muscle cell experiments, we also investigated the direct effects of the
phenolic compounds on the oxidation of pure GSH. Briefly, glutathione (50 mmol/L) was incubated for 10 minutes in phosphate buffer
(50 mmol/L), pH 7.4, both in the absence and in the presence of
phenolic compound, and the levels of glutathione were assayed.
Total lipids were extracted by using a procedure adapted from that of
Folch et al.27 Briefly, methanol (2 mL) containing butylated hydroxytoluene (0.1 mg) was added to the cell suspension (13106
cells). The suspension was mixed with chloroform (4 mL) and kept
for 1 hour under a nitrogen atmosphere on ice in the dark to ensure
complete extraction. After addition of 0.1 mol/L NaCl (1 mL) and
vortex mixing (still under a nitrogen atmosphere), the chloroform
layer was separated by centrifugation (1500g, 5 minutes). The
chloroform was evaporated with a stream of nitrogen, and the lipid
extract was dissolved in 4:3:0.16 (vol/vol) 2-propanol– hexane–water
(0.2 mL). Control experiments demonstrated that the procedure
recovered .95% of cell phospholipids.
HPLC Analysis of Cell Lipids
Lipid extracts were separated with an ammonium acetate gradient
using normal-phase HPLC by slight modification of the method of
Geurts van Kessel et al.28 Briefly, a 5-mm Supelcosil LC-Si column
(4.63250 mm) was used with the following mobile phase flowing at
1 mL/min: solvent A (2-propanol– hexane–water 57:43:1 vol/vol),
solvent B (2-propanol– hexane– 40 mmol/L aqueous ammonium
acetate, pH 6.7), 0- to 3-minute linear gradient from 10% solvent B
to 37% solvent B, 3- to 15-minute isocratic elution at 37% solvent B,
15- to 23-minute linear gradient to 100% solvent B, and 23- to
45-minute isocratic elution at 100% solvent B. A Shimadzu highperformance liquid chromatograph (model LC-600) equipped with
an in-line configuration of fluorescence (model RF-551) and absorbance (model SPD-10A V) detectors was used. The effluent was
monitored at 205 nm to gauge separation of lipids. Fluorescence
emitted by cis-parinaric acid in eluates was monitored fluorometrically at 420-nm emission and 324-nm excitation, and the data were
processed with Shimadzu EZChrom software.
Determination of Lipid Phosphorus in
Lipid Extracts
Lipid phosphorus in lipid extracts was estimated spectrophotometrically at 660 nm as described by Chalvardjian and Rubnicki,29 with
minor modifications. Aliquots of lipid extract were pipetted into test
tubes, and the solvent was evaporated to dryness under a stream of
nitrogen. Perchloric acid (50 mL) was then added to samples, and the
mixtures were incubated for 20 minutes at 170°C to 180°C. After the
tubes were cooled, 0.4 mL of distilled water was added to each tube
followed by 2 mL of sodium molybdate–malachite green reagent
(4.2% sodium molybdate in 5.0N hydrochloric acid/0.2% malachite
green, 1:3 vol/vol). Without delay, 80 mL of 1.5% Tween 20 was
added, mixed immediately to stabilize the color, and read spectrophotometrically at 660 nm.
High-Performance Thin-Layer Chromatograph
Analysis of Cell Lipids
Different classes of phospholipid extracted from smooth muscle cells
were separated by 2-dimensional high-performance thin-layer chromatography on silica G plates (535 cm, Whatman). The plates were
first developed with a solvent system consisting of chloroform:methanol:28% ammonium hydroxide, (65:25:5, vol/vol). After drying, the plates were developed in the second dimension with a
AAPH-Induced Oxidation of Intracellular Thiols
and Glutathione
Determination of Glutathione Concentration in
Smooth Muscle Cells
Glutathione concentration in smooth muscle cells was determined by
estimating the difference in 5,59-dithiobis(2-nitrobenzoic acid)titratable thiols31 in the presence and in the absence of glutathione
peroxidase and cumene hydroperoxide. Smooth muscle cells (53106
cells) were sonicated and then incubated with glutathione peroxidase
(1.94 U/mL), cumene hydroperoxide (333 mmol/L), and deferoxamine mesylate (100 mmol/L) for 30 minutes at 25°C. An aliquot of
cell lysate was then added to 200 mmol/L 5,59-dithiobis(2nitrobenzoic acid), and the precipitated protein was separated by
centrifugation for 5 minutes, 10 000g at 4°C. The absorbance of
supernatant was determined spectrophotometrically at 412 nm.
Growth Studies
[3H]Thymidine incorporation, cell number, and cell migration studies were conducted to investigate the effects of agents on AMVNinduced DNA synthesis, cell proliferation, and cell migration.
Smooth muscle cells were plated at a density of 43104 cells/well in
24-well tissue-culture dishes and allowed to grow to subconfluence
in DMEM/F12 (phenol red–free) medium containing 10% FCS
(steroid free and delipidated) under standard tissue culture conditions. The cells were then growth arrested by feeding with DMEM
(phenol red free) containing 0.4% albumin for 48 hours. For DNA
synthesis, growth was initiated by treating growth-arrested cells for
20 hours with DMEM containing 0.5% FCS and supplemented with
AMVN (1 to 100 mmol/L) and containing or lacking estrone,
2-hydroxyestradiol, 4-hydroxytamoxifen, or droloxifene. After 20
hours of incubation, the treatments were repeated with freshly
prepared solutions but supplemented with [3H]thymidine (1 mCi/mL;
ICN Biomedicals) for an additional 4 hours. The experiments were
terminated by washing the cells twice with Dulbecco’s PBS and
twice with ice-cold trichloroacetic acid (10%). The precipitate was
solubilized in 500 mL of 0.3N NaOH and 0.1% SDS after incubation
at 50°C for 2 hours. Aliquots from 4 wells for each treatment with 10
mL of scintillation fluid were counted in a liquid scintillation
counter. For cell number experiments, smooth muscle cells were
allowed to attach overnight, were growth arrested for 48 hours, were
treated every 24 hours for 4 days, and on day 5 were dislodged and
counted on a Coulter counter.
232
Estrogen Prevents Membrane Peroxidation and Growth
Cell Migration Studies
Modified Boyden chambers (Neuro Probe Inc) were used to evaluate
the effects of AMVN-derived peroxyl radicals on smooth muscle cell
migration as previously described.24 Briefly, confluent monolayers
of smooth muscle cells were growth arrested by feeding with DMEM
supplemented with 0.4% albumin for 48 hours. Growth- arrested
smooth muscle cells were trypsinized, washed, and suspended at a
concentration of 13106 cells/mL in fresh DMEM containing 0.4%
albumin and supplemented with AMVM (50 mmol/L) and containing
or lacking 0.3 to 30 ng/mL estrone, 2-hydroxyestradiol,
4-hydroxytamoxifen, or PMC. Following incubation, the smooth
muscle cells (50 000 cells/50 mL; 6000 cells/mm2) were layered on
the top chamber, and DMEM containing the respective treatments
plus 0.5% FCS was added to the lower chamber. After 6 hours of
incubation, the membranes were removed, the nonmigrated cells on
the top surface wiped, and the migrated cells fixed and stained in Dif
Quick stain (Baxter Scientific Corp). The migrated cells on the lower
surface of the membrane were counted manually at 6 different areas
of the membrane.
Cell Viability Assays
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Trypan blue exclusion, MTT assay, and cell adhesion studies (for
cells treated in suspension) were conducted to confirm cell viability
after various treatments. For the trypan blue exclusion assay, cells
were stained with trypan blue, and the cells taking up the dye were
counted microscopically. For the MTT assay, we used a modified
colorimetric assay based on the selective ability of living cells to
reduce the yellow dye, MTT, to intracellular formazan via mitochondrial enzymes, as described before.32 Briefly, aliquots of cells treated
for different times in suspension with the various experimental
agents were plated in multiwell plates, allowed to attach, and
subsequently treated with MTT (5 mg/mL) under standard tissueculture conditions. In monolayers treated with different agents, MTT
was added directly to the wells after treatment. Following incubation
for 4 hours, the supernatant was aspirated without disruption of the
formazan precipitate, and the intracellular formazan crystals were
dissolved by adding 150 mL of DMSO to each well. The absorbance
was measured at 570 nm using a microplate spectrophotometer. For
cell adhesion viability assays, aliquots of cells treated in suspension
were plated in multiwell plates, and the dead floating cells remaining
after 5 hours of incubation were counted under the microscope. In all
experiments, cells grown in medium and without any treatment
served as controls.
Statistics
All experiments were conducted in triplicate or quadruplicate with 3
to 5 separate cultures. Data are presented as mean6SEM, and
statistical analysis was performed using ANOVA, paired Student’s t
test or Fisher’s least significant difference test as appropriate. A
value of P,0.05 was considered statistically significant.
Results
Incorporation of cis-Parinaric Acid Into Smooth
Muscle Cell Lipids
Figure 1A shows a typical high-performance liquid chromatogram of fluorescence detected in phospholipids extracted from smooth muscle cells incubated in the presence of
hSA-cis-parinaric acid complex. Four phospholipid peaks
were well resolved and identified as phosphatidylethanolamine, phosphatidylserine, phosphatidylcholine, and sphingomyelin. Under the same excitation and emission limits used
above, fluorescent compounds were not detected by HPLC in
extracts from cells incubated with hSA in the absence of
cis-parinaric acid (data not shown).
cis-Parinaric acid incorporation into different membrane
phospholipids in smooth muscle cells was dependent on the
incubation time of these cells with albumin-cis-parinaric acid
Figure 1. Representative HPLC chromatogram showing the fluorescent detection of various phospholipids eluted from total
lipids extracted from 13106 smooth muscle cells incubated with
cis-parinaric acid or cis-parinaric acid plus AMVN (500 mmol/L).
A, Vascular smooth muscle cells labeled with cis-parinaric acid;
control. B, Vascular smooth muscle cells treated with cisparinaric acid plus AMVN. Various phospholipids (phosphatidylethanolamine, PEA; phosphatidylserine, PS; phosphatidylcholine, PC; sphingomyelin, Sph; and free cis-parinaric acid, PnA)
were detected fluorometrically at 324-nm excitation and 420-nm
emission.
complex (Figure 2A). The maximal incorporation of cisparinaric acid into all detected phospholipid classes was
reached within 1 to 2 hours of incubation (Figure 2A).
cis-Parinaric acid was differentially incorporated into
the various phospholipids (phosphatidylcholine.phosphatidylethanolamine.phosphatidylserine.sphingomyelin).
Effect of AMVN on Smooth Muscle Cell
Membrane Lipids
Figure 1B depicts a chromatogram of fluorescence detected
in cis-parinaric acid–acylated lipids extracted from cisparinaric acid–labeled smooth muscle cells incubated in the
presence of AMVN. Incubation of smooth muscle cells with
AMVN (500 mmol/L) induced significant oxidation of cisparinaric acid in all the detected phospholipids (Figure 2B) as
indicated by the reduction in fluorescence. In smooth muscle
cells treated with 50 to 100 mmol/L of AMVN, no oxidation
of cis-parinaric acid–labeled phospholipids could be observed
(Figure 2B). Incubation of cis-parinaric acid–labeled smooth
muscle cells with 250 mmol/L AMVN resulted in a decrease
of cis-parinaric acid fluorescence of phosphatidylserine,
phosphatidylethanolamine, and phosphatidylcholine by 39%,
22%, and 16%, respectively (Figure 2B). The extent of
oxidation of cis-parinaric acid that was incorporated into the
smooth muscle cell phospholipids was dependent on the
duration of incubation of these cells with AMVN (data not
shown). No loss in cell viability was observed in cells treated
for various times and concentrations of AMVN and as
Dubey et al
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Figure 2. A, Time curve for incorporation of cis-parinaric acid
into phospholipids of cultured smooth muscle cells. Cells
(13106/mL) were incubated with 5 mg of cis-parinaric acid/0.5
mg of hSA/mL of L1210 medium for different times at 37°C in
the dark. Lipids from cis-parinaric acid–loaded cells were
extracted and analyzed using normal-phase HPLC with fluorescence detection of various phospholipids (abbreviations as
defined in Figure 1 legend). The incorporation of cis-parinaric
acid in phospholipids was differential, and the relative amounts
of cis-parinaric acid incorporated (ng cis-parinaric acid/mg total
lipid phosphorous) were as follows: phosphatidylcholine.
phosphatidylethanolamine.phosphatidylserine.
sphingomyelin. Each data point represents mean6SEM of 4 or
5 separate experiments conducted in triplicate with separate
cultures. B, Line graph showing the effects of the azo- initiator
of peroxyl radicals, AMVN, on cis-parinaric acid–labeled phospholipids in smooth muscle cells. HPLC analysis shows that
incubation of cis-parinaric acid–loaded cells with AMVN (50 to
500 mmol/L) for 2 hours significantly decreased unoxidized cisparinaric acid content in phospholipids (abbreviations as defined
in Figure 1 legend).Values represent mean6SEM of 4 or 5 separate experiments conducted in triplicate with separate cultures.
*P,0.05 vs control (cells treated with buffer).
assessed by trypan blue exclusion test, MTT assay, and cell
adhesion assay (data not shown).
Effect of Various Estrogens on the
AMVN-Induced Oxidation of cis-Parinaric
Acid–Acylated Phospholipid in Smooth
Muscle Cells
2-Hydroxyestradiol, 4-hydroxytamoxifen, and PMC protected cis-parinaric acid–acylated phospholipids against
AMVN-induced oxidation in a concentration-dependent manner, whereas droloxifene and estrone did not (Figures 3 and
4). The potency of PMC in protecting the various phospholipids was in the following order: phosphatidylethanolamine.phosphatidylserine.phosphatidylcholine'sphingomyelin. Concentrations of 2-hydroxyestradiol and
4-hydroxytamoxifen higher than 250 ng/mL were as effective
as PMC in preventing AMVN-induced oxidations of phosphatidylethanolamine and more effective than PMC in preventing phosphatidylserine, phosphatidylcholine, and sphingomyelin oxidation. However, at lower concentrations (,250
ng/mL), PMC was more effective than 2-hydroxyestradiol
February 5, 1999
233
Figure 3. Comparison of the protective effects of estrone,
2-hydroxyestradiol, and PMC on AMVN-induced oxidation of
cis-parinaric acid-acylated phospholipids in smooth muscle
cells. Cells labeled with cis-parinaric acid were incubated for 2
hours with various concentrations (0.05 to 5 mg/mL) of estrone,
2-hydroxyestradiol, or PMC in the presence and absence of
AMVN (500 mmol/L), and levels of unoxidized cis-parinaric acid
in various phospholipids were subsequently analyzed by HPLC
of total lipid extracts of vascular smooth muscle cells. The data
for the change in levels of unoxidized cis-parinaric acid are presented as percentage of control, with 100% defined as levels of
unoxidized cis-parinaric acid in various fractions of cis-parinaric
acid–labeled cells incubated with vehicle (buffer). Each data
point represents mean6SEM of 4 or 5 separate experiments
conducted in triplicate using separate cultures. *P,0.05 vs
untreated controls or AMVN-treated cells.
and 4-hydroxytamoxifen in protecting all the phospholipids
(Figures 3 and 4). AMVN-induced oxidation of the various
phospholipids was completely blocked (100%) by 2.5 to 5
mg/L per 106 cells of PMC, 2-hydroxyestradiol, and
4-hydroxytamoxifen. The overall efficacy of the various
estrogens and phenolic compounds in protecting against
lipid peroxidation was PMC'2-hydroxyestradiol.4hydroxytamoxifen.droloxifene'estrone (Figures 3 and 4).
In smooth muscle cells pretreated for 20 hours, physiological concentrations (0.3 ng/mL) of 2-hydroxyestradiol selectively decreased AMVN-induced peroxidation of phosphatidylinositol and phosphatidylserine by 34% and 25%,
respectively (Figure 5). In contrast to 2-hydroxyestradiol,
estrone even at high concentrations (30 ng/mL) was a weak
inhibitor of AMVN-induced peroxidation of phosphatidylserine and phosphatidylinositol (Figure 5). Similarly to
2-hydroxyestradiol, 4-hydroxytamoxifen (40 ng/mL) and
PMC (20 ng/mL) selectively protected against AMVNinduced peroxidation of phosphatidylinositol and phosphatidylserine (Figure 5), but not phosphatidylethanolamine and
phosphatidylcholine (data not shown). No loss in cell viability was observed in cells treated for various times and
234
Estrogen Prevents Membrane Peroxidation and Growth
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Figure 4. Comparison of the protective effects of droloxifene,
4-hydroxytamoxifen, and PMC on AMVN-induced oxidation of
cis-parinaric acid–acylated phospholipids in smooth muscle
cells. Cells labeled with cis-parinaric acid were incubated for 2
hours with various concentrations (0.05 to 5 mg/mL) of droloxifene, 4-hydroxytamoxifen, or PMC in the presence and absence
of AMVN (500 mmol/L), and levels of unoxidized cis-parinaric
acid in various phospholipids were subsequently analyzed by
HPLC of total lipid extracts of vascular smooth muscle cells.
The data for the change in levels of unoxidized cis-parinaric
acid are presented as percentage of control, with 100% defined
as levels of unoxidized cis-parinaric acid in various fractions of
cis-parinaric acid–labeled cells incubated with vehicle (buffer).
Each data point represents mean6SEM of 4 or 5 separate
experiments conducted in triplicate using separate cultures.
*P,0.05 vs untreated controls or AMVN-treated cells.
concentrations of the experimental agents and AMVN as
assessed by trypan blue exclusion test, MTT assay, and cell
adhesion assay (data not shown).
Effect of AMVN on the Phospholipid Composition
in Smooth Muscle Cells
High-performance thin-layer chromatographic analysis of total
polar lipids is shown in the Table. Phosphatidylcholine and
phosphatidylethanolamine represented the major fraction of the
total phospholipids in smooth muscle cells and was '50% and
'27% of the total phospholipid content, respectively. Additionally, the other prominent phospholipids in order of their
abundance were as follows: phosphatidylserine.phosphatidylinositol.sphingomyelin.diphosphatidylglycerol.lysophosphatidylcholine.phosphatidic acid. No significant differences in phospholipid distribution or content were detected in
smooth muscle cells following oxidative stress imposed by
incubation of the cells in the presence of AMVN (Table), with
the exception of a slight increase in lysophosphatidylcholine, a
relatively minor phospholipid.
Figure 5. Bar graph comparing the effects of AMVN-induced peroxidation of the phospholipids phosphatidylserine (top panel) and
phosphatidylinositol (bottom panel) in smooth muscle cells pretreated for 20 hours with 2-hydroxyestradiol (2-OH-E; 0.3 ng/mL),
estrone (30 ng/mL), PMC (0.2 ng/mL), or 4-hydroxytamoxifen
(4-OH-T; 40 ng/mL) in the absence of AMVN and subsequently
labeled with cis-parinaric acid and exposed to AMVN (250 mmol/L)
for 3 hours. The levels of unoxidized cis-parinaric acid in various
phospholipids were subsequently analyzed by HPLC in total lipid
extracts of smooth muscle cells. The data for the change in levels
of unoxidized cis-parinaric acid are presented as percentage of
control, with 100% defined as levels of unoxidized cis-parinaric
acid in various fractions of cis-parinaric acid–labeled cells incubated with vehicle (buffer). Each data point represents mean6SEM
of 3 experiments. *P,0.05 vs untreated controls; §P,0.05 vs
AMVN-treated cells.
Effect of Various Estrogens on the AAPH-Induced
Oxidation of Intracellular Thiols and Glutathione
The incubation of smooth muscle cells with AMVN
(500 mmol/L) for 2 hours at 37°C resulted in a mild
oxidation (15% to 17%) of intracellular thiols (glutathione
Effect of AMVN on the Phospholipid Composition of Smooth
Muscle Cells
Percentage of
Total Phospholipids
Phospholipids
Control
AMVN
Phosphatidylcholine
44.262.6
44.162.7
Phosphatidylethanolamine
27.361.5
27.561.6
Phosphatidylinositol
8.760.5
8.460.5
Phosphatidylserine
9.160.7
8.760.6
Sphingomyelin
8.160.5
8.260.6
Diphosphatidylglycerol
1.660.2
1.660.5
Lysophosphatidylcholine
0.560.2
0.960.3*
Phosphatidic acid
0.560.1
0.660.3
Cells were incubated with or without AMVN (500 mmol/L) for 2 hours at 37°C
and various phospholipids separated from the total extracts of phospholipids by
high-performance thin-layer chromatography and quantified. Each value represents percentage of total phospholipids and is mean6SEM from 5 separate
experiments.
*P,0.05 as compared with untreated controls.
Dubey et al
February 5, 1999
235
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Figure 6. Effect of various estrogens (estrone,
2-hydroxyestradiol, droloxifene, and
4-hydroxytamoxifen) and PMC on the AAPHinduced oxidation of intracellular thiols in smooth
muscle cells (top panels) and GSH oxidate rate
(50 mmol/L) in a model system (bottom panels).
Top, Change of thiol content between control and
2-hour incubation of cells in L1210 buffer, pH 7.4,
with 20 mmol/L AAPH at 37°C in the presence of
0.05 to 5 mg/mL of estrone, 2-hydroxyestradiol,
droloxifene, 4-hydroxytamoxifen, or PMC. Each
data point represents mean6SEM of 4 or 5 separate experiments conducted in triplicate using separate cultures. Bottom, The rate of oxidation of
GSH induced by AAPH (20 mmol/L) in the presence and absence of various estrogens and PMC.
GSH (50 mmol/L) in phosphate buffer was incubated for 10 minutes at 37°C with AAPH
(20 mmol/L) in the presence and absence of various agents. Each data point represents
mean6SEM of 4 separate experiments conducted
in triplicate. *P,0.05 vs control (cells incubated
with buffer alone).
plus intracellular thiols; data not shown). The GSH level,
measured as thiol compound after pretreatment of cells with
glutathione-peroxidase and cumene hydroperoxide in the presence of deferoxamine mesylate, was 23.8166.34 nmol/106 cells.
The intracellular glutathione content in smooth muscle cells was
'75% to 85% of total thiol content in smooth muscle cells. The
incubation of smooth muscle cells with AAPH (20 mmol/L) for
2 hours at 37°C resulted in a significant oxidation of the
intracellular thiols and glutathione (Figure 6, top panels). In the
presence of PMC, AAPH-induced oxidation of intracellular
thiols was inhibited in a concentration-dependent manner, suggesting that PMC protects glutathione against oxidation. In
contrast to PMC, 4-hydroxytamoxifen, droloxifene, estrone, and
2-hydroxyestradiol increased the AAPH-induced oxidation rate
of thiols in smooth muscle cells. However, in the absence of
AAPH, these phenolic compounds did not affect the levels of
intracellular thiols and glutathione in the smooth muscle cells
(data not shown). Additionally, when pure glutathione was
incubated in phosphate buffer containing 100 mmol/L deferoxamine mesylate, auto-oxidation of glutathione was not observed.
However, when AAPH (20 mmol/L) was added, the peroxyl
radicals produced by AAPH oxidized glutathione at a rate of
1.45 nmol/min. Moreover, PMC protected glutathione from
AAPH-induced oxidation in a concentration-dependent manner
(Figure 7B). In contrast to the protective effects of PMC, the
estrogens (4-hydroxytamoxifen, droloxifene, estrone, and
2-hydroxyestradiol) did not inhibit AAPH-induced oxidation of
glutathione.
Effects of Various Estrogens on AMVN-Induced
DNA Synthesis, Cell Number, and Cell Migration
Treatment with AMVN stimulated DNA synthesis in a
concentration-dependent manner (P,0.001 versus 0.5% FCS;
Figure 7A). 2-Hydroxyestradiol and 4-hydroxytamoxifen inhibited AMVN-induced DNA synthesis in a concentrationdependent manner. As compared with 2-hydroxyestradiol, estrone was a very weak inhibitor of AMVN-induced DNA
synthesis (Figure 7B). The lowest concentrations of estrone,
2-hydroxyestradiol, and 4-hydroxytamoxifen that inhibited
AMVN-induced DNA synthesis were 0.3 ng/mL for
2-hydroxyestradiol, 30 ng/mL for estrone, and 0.4 ng/mL
for 4-hydroxytamoxifen (Figure 7B). A 50% decrease in
AMVN-induced DNA synthesis in smooth muscle cells was
observed at '3 ng/mL 2-hydroxyestradiol and 40 ng/mL
4-hydroxytamoxifen. In contrast to 2-hydroxyestradiol, estrone
at a concentration of 300 ng/mL inhibited AMVN-induced DNA
synthesis by only 10%. Similarly to the effects on DNA
Figure 7. A, Bar graph showing the concentration-response
relationship for the effects of AMVN-derived peroxyl radicals on
[3H]thymidine incorporation. Each data point represents
mean6SEM from 3 experiments conducted in quadruplicate.
*P,0.05 vs control (C; vascular smooth muscle cells treated
with 0.5% FCS). B, Line graph showing the concentrationresponse relationship for the inhibitory effects of
2-hydroxyestradiol, estrone, and 4-hydroxytamoxifen on
50 mmol/L AMVN–induced DNA synthesis. Results are
expressed as percentage of control, with 100% defined as thymidine incorporation in the presence of 50 mmol/L AMVN. Each
data point represents mean6SEM from 3 experiments, each
conducted in quadruplicate. *P,0.05 vs AMVN.
236
Estrogen Prevents Membrane Peroxidation and Growth
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Figure 8. A, Bar graph showing the effects of AMVN-derived
peroxyl radical on cell number. Treatment of smooth muscle
cells for 4 days with 50 mmol/L AMVN–induced cell proliferation.
C indicates control. Accompanying line graph depicts the
concentration-response relationship for the inhibitory effects of
2-hydroxyestradiol, estrone, and 4-hydroxytamoxifen on
50 mmol/L AMVN–induced cell proliferation. Results are
expressed as percentage of control, with 100% defined as cell
number on day 4 in the presence of 50 mmol/L AMVN. Each
data point represents mean6SEM from 3 experiments, each
conducted in quadruplicate. *P,0.05 vs smooth muscle cells
treated with 0.5% FCS alone or 0.5% FCS plus AMVN. B, Bar
graph showing the effects of AMVN (50 mmol/L) on 0.5% FCSinduced smooth muscle cell migration. Serum-induced migration
was enhanced in smooth muscle cells pretreated with
50 mmol/L AMVN (mean6SEM of 6 observations at 3200 highpower field from 3 separate experiments). C indicates control.
Accompanying line graph shows the inhibitory effects of
2-hydroxyestradiol, estrone, and 4-hydroxytamoxifen on AMVNinduced migration of smooth muscle cells. Results are
expressed as percentage of control, with 100% defined as number of AMVN-treated smooth muscle cells migrating in response
to 0.5% FCS; numbers in 4 separate experiments were the following: 243610, 142617, 293616, and 318612. Each data
point represents mean6SEM from 4 experiments. *P,0.05 vs
smooth muscle cells treated with 0.5% FCS alone or 0.5% FCS
plus AMVN.
synthesis, treatment with AMVN for 4 days increased cell
number, and these proliferative effects were significantly inhibited by 2-hydroxyestradiol and 4-hydroxytamoxifen, but not by
estrone (Figure 8A).
FCS stimulated migration of smooth muscle cells, and this
effect was significantly enhanced in smooth muscle cells
treated with AMVN (Figure 8B). In cells stimulated with FCS
plus AMVN, 2-hydroxyestradiol and 4-hydroxytamoxifen
inhibited migration in a concentration-dependent manner
(Figure 8B), whereas estrone did not inhibit migration. The
lowest concentrations of 2-hydroxyestradiol and
4-hydroxytamoxifen that inhibited cell migration was 0.3
ng/mL for 2-hydroxyestradiol and 0.4 ng/mL for
4-hydroxytamoxifen (Figure 8B).
Discussion
The present study demonstrates that aortic smooth muscle
cells labeled with cis-parinaric acid provide a sensitive and
quantitative method to assess the effects of drugs on lipid
peroxidation. Using this technique, this study establishes that
AMVN-induced peroxidation of various smooth muscle cell
membrane phospholipids is not prevented by estrone, an
estrogen-receptor ligand, but is completely blocked by
2-hydroxyestradiol, a major metabolite of 17b-estradiol
that does not have a high affinity for the estrogen receptor.33,34 Moreover, 4-hydroxytamoxifen (a metabolite of tamoxifen with high affinity for estrogen receptors) effectively
blocks peroxidation of smooth muscle cell phospholipids,8
while droloxifene, a tamoxifen analog known to be an
estrogen-receptor ligand,35 does not prevent peroxidation.
Both 2-hydroxyestradiol and 4-hydroxytamoxifen are as effective as PMC in protecting membrane phospholipids
against peroxidation. Moreover, 2-hydroxyestradiol and
4-hydroxytamoxifen inhibit peroxyl radical–induced proliferation and migration of smooth muscle cells. In contrast to the
effects on lipid peroxidation and cell growth,
2-hydroxyestradiol and 4-hydroxytamoxifen do not prevent
AAPH-induced oxidation of cellular thiols/glutathione in
smooth muscle cells. These studies provide the first evidence
that the cardioprotective effects of estrogen and tamoxifen
may in part be due to the direct antioxidant effects of their
major metabolites on membrane phospholipids of smooth
muscle cells.
Because our aim was to evaluate oxidative damage in cells
membranes, we required a sensitive method for detecting
oxidation of phospholipids. Thus, we selected the cisparinaric acid/AMVN paradigm. In this regard, our results
indicate that cis-parinaric acid covalently labels various
phospholipids in smooth muscle cells in a differential and
time-dependent fashion and that cis-parinaric acid fluorescence in the various phospholipids is uniformly reduced in
smooth muscle cells treated with AMVN. These results
indicate that smooth muscle cells can be labeled with cisparinaric acid and that AMVN-derived peroxyl radicals cause
peroxidation of a broad array of lipid components in the cell
membrane. Even though cis-parinaric acid–labeled phospholipids constitute ,1% of the total lipid content,36 we can
detect AMVN-induced peroxidation in cis-parinaric acid–
labeled smooth muscle cells, as well as protection against
AMVN-induced peroxidation by PMC, a potent antioxidant.
The actual content of various membrane phospholipids analyzed using conventional methods did not vary in control and
AMVN-treated smooth muscle cells, indicating that decreases
in unoxidized cis-parinaric acid in labeled smooth muscle
cells were not due to loss of cis-parinaric acid–labeled
phospholipids, but rather to the change in the oxidization state
of incorporated cis-parinaric acid. Moreover, the finding that
cell viability was not reduced in smooth muscle cells treated
with AMVN suggests that the effects of AMVN on lipid
peroxidation were not due to cell death. Together, these
findings strongly suggest that cis-parinaric acid–labeled
smooth muscle cells provide an excellent in vitro system to
evaluate the effects of both oxidative stress-induced lipid
peroxidation and the antioxidant capabilities of chemicals and
new therapeutic drugs. In the present study, we found that
AMVN-induced peroxidation of various membrane phospholipids and AMVN-induced cell growth are not prevented by
estrone, but are completely blocked by 2-hydroxyestradiol.
Dubey et al
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
This finding suggests that the cardioprotective effects of
17b-estradiol may, in part, be due to the direct antioxidant
effects of its metabolite on smooth muscle cells. Differences in the antioxidant capabilities of estrone and
2-hydroxyestradiol may largely be due to the chemical
nature of the molecule. Indeed, 2-hydroxyestradiol is a
catechol estrogen and is a more potent antioxidant than
17b-estradiol.6,33 It is also possible that, similar to 17bestradiol, the metabolism of estrone to 2-hydroxyestrone is
necessary to induce its antioxidant effects. Indeed, a recent
study has shown that 2-hydroxyestrone is several times
more potent than estrone in protecting against lipid peroxidation.6 Moreover, as compared with 17b-estradiol,
2-hydroxyestradiol is several times more potent in inhibiting DNA synthesis in smooth muscle cells37 and proliferation as well as extracellular matrix synthesis in cardiac
fibroblast.38
Whether the cardioprotective effects of estrogens are
receptor dependent is under intense debate. Recent
studies suggest that the cardioprotective effects of estrogen(s) are non–receptor mediated39; however, the mechanisms involved remain unclear. Our finding that
4-hydroxytamoxifen, a phenolic metabolite of tamoxifen
and an estrogen-receptor ligand, is as potent as
2-hydroxyestradiol in protecting phospholipids against
AMVN-induced peroxidation and our finding that estrone
and droloxifene, which bind to estrogen receptors with
high affinity, are ineffective in preventing peroxidation,
suggest that the non–receptor-mediated cardioprotective
effects of estrogens may be due to direct antioxidant
effects of their metabolites on smooth muscle cells. This
notion is further supported by our observation that, similarly to 2-hydroxyestradiol, 4-hydroxytamoxifen inhibited
AMVN-induced growth of smooth muscle cells. Nonetheless, further experiments will be required to determine
whether the observed protective effects of estrogen and
tamoxifen metabolites are greater in females and whether
the effects are blocked by specific estrogen receptor
antagonists.
To evaluate the antioxidant potency of various estrogens, we compared their effects with PMC, a homolog of
the natural membrane antioxidant, and growth inhibitor
vitamin E,40,41 a very potent antioxidant. An important
observation in the present study is that 2-hydroxyestradiol
and 4-hydroxytamoxifen are as effective as PMC in
preventing lipid peroxidation. It is well documented that
4-hydroxytamoxifen is severalfold more potent than tamoxifen in preventing LDL oxidation, iron-induced
microsomal peroxidation, and cardiovascular disease.8
Moreover, it has been suggested that the cardioprotective
effects of tamoxifen are largely mediated via generation
of 4-hydroxytamoxifen. This is in agreement with our
observation that low concentrations (1 nmol/L) of
4-hydroxytamoxifen inhibited AMVN-induced growth of
smooth muscle cells. With regard to 2-hydroxyestradiol,
no studies have been conducted to compare its potency
with that of 17b-estradiol in inducing cardioprotective
effects in vivo. However, numerous studies suggest that
2-hydroxyestradiol is the major metabolite of 17b-estradiol and
February 5, 1999
237
that some biological effects of estrogens are mediated in part
through their metabolites.33,34 As compared with 17b-estradiol
and estrone, their 2-hydroxymetabolites (2-hydroxyestradiol and
2-hydroxyestrone) are severalfold more effective in preventing
lipid peroxidation in liposomes.6,10,13,33 Hence, similarly to
tamoxifen,8 estrogens may largely induce their cardioprotective
effects via generation of the potent antioxidant metabolites. This
hypothesis is further supported by our observation that low
concentrations of 2-hydroxyestradiol, but not estrone, inhibited
free radical–induced DNA synthesis in smooth muscle cells.
The cardioprotective effects of several drugs with antioxidant effects are thought to be mediated via modulation
of the endogenous free radical scavenging system, including the glutathione reductase system.42 Hence, the possibility that estrone, 2-hydroxyestradiol, droloxifene, and
4-hydroxytamoxifen may also induce their cardioprotective effects by preventing peroxyl radical–induced oxidation of intracellular thiols in smooth muscle cells was also
investigated. In smooth muscle cells treated with AAPH,
the intracellular thiol/GSH is oxidized rapidly. In contrast
to the effects on lipid peroxidation, 4-hydroxytamoxifen
and 2-hydroxyestradiol did not prevent AAPH-induced
oxidation of smooth muscle cell thiol/glutathione. Thus,
the antioxidant effects of 2-hydroxyestradiol and
4-hydroxytamoxifen are exerted within the cell membrane,
rather than in the cytoplasm.
Do estrogens prevent lipid peroxidation and free radical–induced growth in vivo? The observation that physiological concentrations (0.3 ng/mL) of 2-hydroxyestradiol
inhibited free radical–induced growth and migration of
smooth muscle cells suggests that 2-hydroxyestradiol may
importantly contribute to the cardioprotective effects of
estrogen in vivo. Moreover, the finding that pretreatment
of cells for 20 hours with physiological concentrations of
2-hydroxyestradiol, but not estrone, selectively protects
phosphatidylinositol and phosphatidylserine against
AMVN-induced peroxidation suggests that estradiol metabolites may protect against peroxyl radical–induced growth via
mechanisms linked to these phospholipids. In this regard, it is
important to note that cell viability, as assessed by trypan
blue exclusion, MTT assay, and a cell adhesion assay,
was not reduced by prolonged treatment with AMVN,
2-hydroxyestradiol, 4-hydroxytamoxifen, or estrone. Therefore,
cell death cannot account for the reduced AMVN-induced
peroxidation by 2-hydroxyestradiol. Finally, it should be mentioned that plasma levels of 17b-estradiol do not reflect the true
physiological levels of antioxidant estrogens in the cell membrane, since in addition to 17b-estradiol, several other estrogens
and their metabolites are present in vivo,33 and estrogen metabolites concentrate in cell membranes.43– 45
Regarding the mechanisms mediating oxidation-induced
cell growth, peroxidation of phospholipids has been shown
to activate the extracellular signal–regulated protein kinases (ERK1 and ERK2), induce the expression of c-fos
and c-jun oncogene proteins, increase activator protein-1
DNA binding activity, and induce nuclear factor-kb.43– 48
Moreover, the acidic membrane phospholipids phosphatidylinositol and phosphatidylserine are known to selectively activate protein kinase C activity, whereas phos-
238
Estrogen Prevents Membrane Peroxidation and Growth
phatidylethanolamine and phosphatidylcholine do not.45
Moreover, oxidized metabolites of phosphatidylinositol
have been shown to induced cell migration. Hence, it is
possible that 2-hydroxyestradiol and 4-hydroxytamoxifen
prevent smooth muscle cell growth by regulating protein
kinase C activity by selectively protecting peroxidation of
phosphatidylinositol and phosphatidylserine. However,
further detailed studies will be required to clarify the
mechanisms by which peroxidation of phospholipids enhances cell growth.
In conclusion, our findings suggest that 2-hydroxyestradiol
and 4-hydroxytamoxifen may be more potent and clinically
important than their parent compounds in preventing lipid
peroxidation, growth of smooth muscle cells, and vaso-occlusive
disorders.
Acknowledgments
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
This work was supported by grants from the National Institutes of
Health (HL 55314, HL 35909, and GM 47124), the Swiss National
Science Foundation (32-54172.98), and in part by International
Neurological Science Fellowship Program F05 NS10669, administered by National Institute of Neurological Disorders and Stroke,
National Institutes of Health, in collaboration with Unit of Neuroscience, Division of Mental Health and Substance Abuse, World
Health Organization.
References
1. Lerner DJ, Kannel WB. Patterns of coronary heart disease morbidity and
mortality in the sexes: a 26-year follow-up of the Framingham population.
Am Heart J. 1986;111:383–390.
2. Wenger NK, Speroff L, Packard B. Cardiovascular health and disease in
women. N Engl J Med. 1993;329:247–256.
3. Stampfer MJ, Colditz GA. Estrogen replacement therapy and coronary
heart disease: a quantitative assessment of the epidemiologic evidence.
Prev Med. 1991;20:47– 63.
4. Notelovitz M. Estrogen replacement therapy: indications, contraindications, and agent selection. Am J Obstet Gynecol. 1989;161:1832–1841.
5. American Heart Association. Heart and Stroke Facts. Dallas, TX:
American Heart Association; 1992.
6. Niki E, Nakano M. Estrogens as antioxidants. Methods Enzymol. 1990;
186:330 –336.
7. MacDonald CC, Stewart HJ. Fatal myocardial infarction in the Scottish
adjuvant tamoxifen trial. BMJ. 1991;303:435– 437.
8. Wiseman H, Paganga G, Rice-Evans C, Halliwell B. Protective actions of
tamoxifen and 4-OH-tamoxifen against oxidative damage to human lowdensity lipoproteins: a mechanism accounting for the cardioprotective
action of tamoxifen. Biochem J. 1993;292:635– 638.
9. Shwaery GT, Vita JA, Keaney JF. Antioxidant protection of LDL by
physiological concentrations of 17b-estradiol: requirement for estradiol
modification. Circulation. 1997;95:1378 –1385.
10. Rifici VA, Khachadurian AK. The inhibition of low-density lipoprotein
oxidation by 17b-estradiol. Metabolism. 1992;41:1110 –1114.
11. Maziere C, Auclair M, Ronveaux MF, Salmon S, Santus R, Maziere J-C.
Estrogens inhibit copper and cell-mediated modification of low density
lipoprotein. Atherosclerosis. 1991;89:175–182.
12. Keaney JF, Shwaery GT, Xu A, Nicolosi RJ, Loscalzo J, Foxall TL, Vita
JA. 17b-Estradiol preserves endothelial vasodilator function and limits
low-density lipoprotein oxidation in hypercholesterolemic swine. Circulation. 1994;89:2251–2259.
13. Tang M, Abplanalp W, Ayres S, Subbiah MTR. Superior and distinct
antioxidant effects of selected estrogen metabolites on lipid peroxidation.
Metabolism. 1996;45:411– 414.
14. Oparil S, Levine RL, Chen YF. Sex hormones and the vasculature. In:
Sowers JR, Walsh M, eds. Endocrinology of the Vasculature. Totwa, NJ:
Humana Press, Inc; 1996:225–237.
15. Krasinski K, Spyridopoulos I, Asahara T, van der Zee R, Isner JM,
Losordo DW. Estradiol accelerates functional endothelial recovery after
arterial injury. Circulation. 1997;95:1768 –1772.
16. Oparil S, Levine RL, Chen S-J, Durand J, Chen YF. Sexually dimorphic
response of the balloon-injured rat carotid artery to hormone treatment.
Circulation. 1997;95:1301–1307.
17. Foegh M, Asotra S, Howell M, Ramwell P. Estradiol inhibition of arterial
neointimal hyperplasia after balloon injury. J Vasc Surg. 1994;14:
722–726.
18. Kehrer JP. Free radicals as mediators of tissue injury and disease. Crit
Rev Toxicol. 1993;23:21– 48.
19. Fabisiak JP, Kagan VE, Ritov VB, Johnson DE, Lazo JS. Bcl-2 inhibits
selective oxidation and externalization of phosphatidylserine during
paraquat-induced apoptosis. Am J Physiol. 1997;272:C675–C684.
20. Keaney JF Jr, Vita JA. Atherosclerosis, oxidative stress, and antioxidant
protection in endothelium-derived relaxing factor action. Prog Cardiovasc Dis. 1995;38:129 –154.
21. Alexander RW. Hypertension and the pathogenesis of atherosclerosis:
oxidative stress and the mediation of arterial inflammatory response: a
new perspective. Hypertension. 1995;25:155–161.
22. Ritov VB, Banni S, Yalowich JC, Day BW, Claycomp HG, Corongiu FP,
Kagan VE. Non-random peroxidation of different classes of membrane
phospholipids in live cells detected by metabolically integrated cisparinaric acid. Biochim Biophys Acta. 1996;1283:127–140.
23. Krainev AG, Bigelow DJ. Comparison of 2,29-azobis(2-amidinopropane)
hydrochloride (AAPH) and 2,29-azobis(2,4-dimethylvaleronitrile)
(AMVN) as free radical initiators: a spin-trapping study. J Chem Soc
Perkin Trans. 1996;2:747–754.
24. Dubey RK, Jackson EK Luscher TF. Nitric oxide inhibits angiotensin
II-induced migration of rat aortic smooth muscle cell: role of cyclicnucleotides and angiotensin 1 receptors. J Clin Invest. 1995;96:
141–149.
25. Dubey RK, Roy A, Overbeck HW. Culture of renal arteriolar smooth
muscle cells: mitogenic responses to angiotensin II. Circ Res. 1992;71:
1143–1152.
26. Niki E. Free radical initiators as source of water- or lipid-soluble peroxyl
radicals. Methods Enzymol. 1990;186:100 –108.
27. Folch J, Lees M, Sloan-Stanley GH. A simple method for isolation and
purification of total lipids from animal tissues. J Biol Chem. 1957;226:
497–509.
28. Geurts van Kessel WSM, Hax WMA, Demel RA, DeGier J. High
performance liquid chromatographic separation and direct ultraviolet
detection of phospholipids. Biochem Biophys Acta. 1977;486:
524 –530.
29. Chalvardjian A, Rubnicki E. Determination of lipid phosphorous in the
nanomolar range. Anal Biochem. 1996;36:225–226.
30. Arduini A, Tyurin V, Tyuruna Y, Arrigoni-Martelli E, Molajoni F,
Dottori S, Federici G. Acyl-trafficking in membrane phospholipid fatty
acid turnover: the transfer of fatty acid from acyl-L-carnitine pool to
membrane phospholipids in intact human erythrocytes. Biochem Biophys
Res Commun. 1992;187:353–358.
31. Ellman GL. Tissue sulfhydryl groups. Arch Biochem Biophys. 1959;82:
302–310.
32. Carmichael J, DeGraff WG, Gazdar AF, Minna JD, Mitchell JB.
Evaluation of a tetrazolium-based semiautomated colorimetric assay:
assessment of chemosensitivity testing. Cancer Res. 1987;47:
936 –942.
33. Gelbke HP, Ball P, Knuppen R. 2-Hydroxyoestrogens: chemistry, biogenesis, metabolism and physiological significance. In: Briggs MH,
Christie GA, eds. Advances in Steroid Biochemistry and Pharmacology.
Vol 6. New York, NY: Academic Press; 1977:81–154.
34. Adlercreutz H, Gorbach SL, Goldin BR, Woods MN, Dwyer JT,
Hämäläinen E. Estrogen metabolism and excretion in Oriental and Caucasian women. J Natl Cancer Inst. 1994;86:1076 –1082.
35. Ke HZ, Simmons HA, Pirie CM, Crawford DT, Thompson DD.
Droloxifene, a new estrogen antagonist/agonist, prevents bone loss in
ovariectomized rats. Endocrinology. 1995;136:2435–2441.
36. van den Berg JJM, Kuypers FA, Lubin BH, Roelofsen B, Op den Kamp
JAF. Direct and continuous measurement of hydroperoxide-induced oxidative stress on the membrane of intact erythrocytes. Free Radical Biol
Med. 1991;11:255–261.
37. Nishigaki I, Sasaguri Y, Yagi K. Anti-proliferative effect of
2-methoxyestradiol on cultured smooth muscle cells from rabbit aorta.
Atherosclerosis. 1995;113:167–170.
38. Dubey RK, Gillespie DG, Jackson EK, Keller PJ. 17b-Estradiol, its
metabolites and progesterone inhibit cardiac fibroblast growth. Hypertension 1998;31(1 pt 2):522–528.
Dubey et al
39. Isfrati MD, Karas RH, Aronovitz M, Kim S, Sullivan TR Jr, Lubahn DB,
O’Donell TF Jr, Korach KS, Mendelsohn ME. Estrogen inhibits the
vascular injury response in estrogen receptor a-deficient mice. Nat Med.
1997;3:545–548.
40. Tyurina YY, Tyurin VA, Yalowich JC, Quinn PJ, Claycamp HG,
Schor NF, Pitt BR, Kagan VE. Phenoxyl radicals of etoposide (VP-16)
can directly oxidize intracellular thiols: protective versus damaging
effects of phenolic antioxidants. Toxicol Appl Pharmacol. 1995;131:
277–288.
41. Boscoboinik DO, Chatelain E, Bartoli GM, Stauble B, Azzi A. Inhibition
of protein kinase C activity and vascular smooth muscle cell growth by
D-alpha-tocopherol. Biochim Biophys Acta. 1994;122:418 – 426.
42. Meister A, Anderson ME. Glutathione. Annu Rev Biochem. 1983;52:
711–760.
43. Rao GN, Alexander RW, Runge MS. Linoleic acid and its metabolites,
hydroperoxyocta-decadienoic acids, stimulate c-Fos, c-Jun and c-Myc
mRNA expression, mitogen-activated protein kinase activation, and
44.
45.
46.
47.
48.
February 5, 1999
239
growth in rat aortic smooth muscle cells. J Clin Invest. 1995;96:
842– 847.
Ruef J, Rao GN, Li F, Bode C, Patterson C, Bhatnagar A, Runge MS.
Induction of rat aortic smooth muscle cell growth by lipid peroxidation product 4-hydroxy-2-nonenal. Circulation. 1998;97:
1071–1078.
Nakanishi H, Brewer KA, Exton JH. Activation of the zeta isozyme of
protein kinase C by phosphatidylinositol 3,4,5-triphosphate. J Biol Chem.
1993;268:13–16.
Sen CK, Packer L. Antioxidant and redox regulation of gene transcription. FASEB J. 1996;10:709 –720.
Lander HM. An essential role for free radicals and derived species in
signal transduction. FASEB J. 1991;11:118 –124.
Larner JM, Pahuja SL, Shackleton CH, McMurray WJ, Giordano G,
Hochberg RB. The isolation and characterization of estradiol-fatty acid
esters in human ovarian follicular fluid: identification of an endogenous
long-lived and potent family of estrogens. J Biol Chem.
1993;268:13893–13899.
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Estrogen and Tamoxifen Metabolites Protect Smooth Muscle Cell Membrane
Phospholipids Against Peroxidation and Inhibit Cell Growth
Raghvendra K. Dubey, Yulia Y. Tyurina, Vladimir A. Tyurin, Delbert G. Gillespie, Robert A.
Branch, Edwin K. Jackson and Valerian E. Kagan
Circ Res. 1999;84:229-239
doi: 10.1161/01.RES.84.2.229
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1999 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/84/2/229
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/