Mechanisms involved in the protective effect
of estradiol-17b on lipid peroxidation and DNA damage
STACEY AYRES,1 WILLIAM ABPLANALP,2 JAMES H. LIU,3 AND M. T. RAVI SUBBIAH2
Graduate Studies Program and 2Division of Endocrinology
and Metabolism, Department of Internal Medicine, and 3Department of Obstetrics
and Gynecology, University of Cincinnati Medical Center, Cincinnati, Ohio 45267-0540
1Interdisciplinary
reactive oxygen species
THERE IS CONSIDERABLE EVIDENCE that oxidative DNA
damage mediated by reactive oxygen radicals and lipid
peroxidation may play a role in carcinogenesis and
atherogenesis (2). Reactive oxygen species, especially
in the presence of transition metals, are potent stimulators of lipid peroxidation and DNA damage (10). A
variety of antioxidant defenses, including enzymes,
have evolved to protect against reactive oxygen species,
but these defenses are not completely efficient or effective. As a result, there is an accumulation of oxidative
damage to DNA, lipids, and proteins during the aging
process. Although the chemical mechanisms are still
not fully understood, oxidized lipids and reactive oxygen species have been shown to damage DNA (3, 27);
thus attempts to prevent these damaging effects are
essential. Estrogens have been shown to be powerful
antioxidants, effectively preventing lipid peroxidation
(4, 14, 23). In particular, estrogen decreases oxidative
modification of low-density lipoproteins (LDL), both in
E1002
vitro and in vivo (9, 10, 14, 19, 21, 23). It is postulated
that this antioxidant activity is considered to be one
mechanism by which estrogen confers cardioprotection
(7).
The chemical structure of estrogen allows for donation of a H1 atom to a peroxyl radical. This property of
estrogen allows free radical scavenging and may exert
its effect by interfering early or during the propagation
phase of lipid peroxidation. Recently, we presented
evidence that estradiol-17b or estrogen (E2 ) was as
effective as vitamin E in preventing LDL peroxidation
and cholesterol oxidation (4). We postulated that E2
may act by regenerating or maintaining endogenous
antioxidants in lipoproteins and thus delaying the
appearance of peroxidized products. Our studies suggested that the protection or regeneration of antioxidant vitamin content in the lipoprotein is not the
mechanism of E2 action, because E2 had no effect on the
content of antioxidants in lipoproteins, and no protective effect was observed after reisolation of LDL and
high-density lipoproteins (HDL) incubated with E2.
These experiments suggest that E2 may be exerting its
effect by interfering early in the peroxidation process,
perhaps by modifying the reaction pathway responsible
for the propagation phase in the generation of reactive
oxygen species (ROS), such as superoxide, H2O2, or ȮH
radicals.
In addition to its protective role in preventing lipid
peroxidation, E2 has been shown to protect against
DNA damage. Tang and Subbiah (24), utilizing a chemical model system consisting of OX-174 RF1 DNA, a
supercoiled DNA, demonstrated that E2 protected
against DNA damage induced by hydrogen peroxide
and arachidonic acid, as indicated by DNA strand
breakage. However, the precise mechanism of E2’s
protective effect on DNA damage was not clear.
To further evaluate the antioxidant property and
mechanism responsible for the protective effect of E2 on
lipid peroxidation and DNA damage, studies were
conducted 1) to compare the effect of E2 with that of
selective scavengers of ROS in terms of lipid peroxidation and DNA damage and 2) to determine the direct
effect of E2 on the production of individual ROS. Our
studies suggest that E2 might offer protection from lipid
peroxidation and DNA damage by inhibiting the formation of superoxide radical and also by interfering with
the oxidative chain propagation leading to lipid peroxidation.
MATERIALS AND METHODS
AAPH (2,28-azobis[2-amidinopropane]dihydrochloride), cupric sulfate, hydrogen peroxide (30% solution), EDTA,
0193-1849/98 $5.00 Copyright r 1998 the American Physiological Society
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
Ayres, Stacey, William Abplanalp, James H. Liu, and
M. T. Ravi Subbiah. Mechanisms involved in the protective
effect of estradiol-17b on lipid peroxidation and DNA damage.
Am. J. Physiol. 274 (Endocrinol. Metab. 37): E1002–E1008,
1998.—Previous studies from our laboratory have shown that
estrogens can protect against lipoprotein peroxidation and
DNA damage. In this study, the mechanism of estradiol-17b
(E2 ) action was investigated by comparing E2 with selective
scavengers of reactive oxygen species (ROS) in terms of
inhibition of 1) human low-density lipoprotein (LDL) peroxidation (measured by the diene conjugation method) and 2)
DNA damage (measured by the formation of strand breaks in
supercoiled OX-174 RFI DNA). In addition, the direct effect of
E2 on the generation of individual ROS was also measured.
By use of ROS scavengers, it was determined that lipoprotein
peroxidation was predominantly due to superoxide (39%),
with some contributions from hydrogen peroxide (23%) and
peroxy (38%) radicals. E2 was a more effective inhibitor of
peroxidation than all the ROS scavengers combined. In DNA
damage, scavengers of hydrogen peroxide, hydroxyl, and
superoxide radical offered significant protection (49–65%). E2
alone offered a similar degree of protection, and no additional
effect was evident when it was combined with ROS scavengers. E2 caused a significant reduction (37%) in the production of superoxide radical by bovine heart endothelial cells in
culture but had no effect on the formation of either hydrogen
peroxide or hydroxyl radicals. These studies show that 1) the
protection offered by E2 in terms of lipid peroxidation could be
due to its ability to inhibit generation of superoxide radical
and prevent further chain propagation, and 2) in DNA
damage protection, E2 mainly appears to inhibit chain propagation.
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
ethidium bromide, OX-174 RFI DNA, estradiol-17b, atocopherol, b-carotene, superoxide dismutase (SOD), catalase, mannitol, and sodium azide were purchased from Sigma
Chemical (St. Louis, MO). Additional chemicals and reagents
used in detecting formation of ROS include salicylic acid
(Aldrich Chemical, Milwaukee, WI); 2,3-, 2,4-, and 2,5dihydroxybenzoic acid (DHBA; Aldrich); HPLC-grade diethyl
ether (Aldrich); HPLC-grade methanol (Aldrich); and menadione (1,2,3,4-tetrahydro-2-methyl-1,4-dioxo-2-naphthalenesulfonic acid sodium salt; Sigma Chemical). Bovine heart microvessel endothelial cells (BHMEC) were obtained from
Gensia (San Diego, CA).
Isolation of Plasma and Lipoproteins
Assessment of Lipid Peroxidation
Effect of E2, vitamins, and free radical scavengers on lipid
peroxidation was evaluated by the diene conjugation method.
The oxidation of lipoproteins was followed by measuring the
absorbency at 234 nm resulting from the formation of conjugated dienes from unsaturated fatty acids, as described by
Esterbauer et al. (8). The time profile of the 234 nm absorption curve shows three distinct phases: lag phase (absorption
does not increase or only slightly increases, indicating that
the lipoprotein resists oxidation), propagation phase (absorption rapidly increases to a maximum value, indicative of the
chain reaction lipoprotein peroxidation process), and decomposition phase (absorption decreases again as conjugated
dienes slowly decrease and decomposition reactions predominate) (8). The formation of conjugated dienes was measured
by incubating 60 µg LDL with 5.5 mM/l AAPH in 0.985 ml
PBS medium. The absorbency at 234 nm was measured
continuously in a thermostat-controlled (37°C) computerized
Beckman DU-64 spectrophotometer equipped with a sixposition automatic sampler changer. The increase in 234 nm
absorption was recorded every 5 min during a 3-h period. The
susceptibility of the lipoproteins to oxidation was assessed on
the basis of lag time (min), the time interval between the
addition of AAPH and the intercept of the slope of the
absorbance curve (8). The maximal amount of dienes formed
was determined as described by Esterbauer et al.
In most studies of lipid peroxidation, an E2 concentration of
54 µM was used. This concentration was chosen because at
this level, significant inhibition of lipid peroxidation was
noted in our previous study (4). In some experiments, ROS
scavengers SOD (0.1 mg/ml), catalase (0.1 mg/ml), and mannitol (5 mM) were added independently or conjointly to the
reaction mixture to evaluate the effect of selective scavengers
on ROS generation.
Assessment of DNA Damage
OX-174 RFI DNA contains ,85–90% double-stranded,
covalently closed, supercoiled DNA molecules, and 10–15%
double-stranded, open, circular DNA with no linear DNA
molecules. DNA strand breaks were determined by measuring the conversion of double-stranded supercoiled OX-174
RFI DNA (form A) to double-stranded open circular DNA
(form B) and linear DNA (form C), as described by Li and
Trush (11). The experiments were performed in 1.5-ml Eppendorf tubes. DNA (0.2 µg) was incubated with H2O2 (15 µM),
copper ion (33 µM), and estrogens (7–36 3 1026 M) in PBS
(final volume 30 µl) at 37°C for 30 min. After incubation, the
samples were loaded in a 1% agarose gel containing 40 mM
Tris-acetate buffer with 1 mM EDTA and were subjected to
electrophoresis in the same buffer for 2 h. The gels were
stained with a solution of ethidium bromide and destained in
distilled water. The gels were photographed, and quantification of DNA bands was performed by band densitometer
tracing with a computer-assisted imaging system. The percentage of various DNA forms was calculated by optical density
values of each band.
In most studies of DNA damage, the E2 concentration
ranged from 7 to 36 µM. This range was chosen because 1) at
these levels significant inhibition of DNA damage was noted
in our previous study (24), and 2) this range allowed manipulations for the study of mechanisms. In those instances when
significant changes were not evident, concentrations were
changed accordingly. To evaluate the effect of selective scavengers on ROS generation and DNA damage, in some experiments ROS scavengers SOD (0.1 mg/ml), catalase (0.1 mg/
ml), and mannitol (5 mM) were added independently or
conjointly to the reaction mixture (5).
Effect of E2 on Formation of ROS
Superoxide detection by cytochrome c reduction. BHMEC
were grown in minimal essential medium (MEM) supplemented with 10% horse serum and penicillin or streptomycin
and were incubated for 30 min at 37°C with menadione (200
µM) under 5% CO2. SOD-inhibitable cytochrome c reduction
experiments as described by McCord and Fridovich (15) were
carried out: cells in tissue culture were washed three times
with buffer and incubated in Krebs bicarbonate-buffered
perfusate solution [(in mM) 110 NaCl, 2.6 KCl, 1.2 MgSO4, 1.2
KH2PO4, 25 NaHCO3, 25 HEPES, and 11 glucose] containing
20 µM ferricytochrome c. Pairs of cultures dishes, one with
and the other without SOD, were exposed to the experimental
conditions at 37°C. The range of E2 concentration used was
0–100 µM. The cells were saved for determination of protein
content, the supernatants were collected, and the absorbance
was determined at 550 nm. Subsequently, the SOD-inhibitable reduction of cytochrome c was calculated using a
molar extinction coefficient of 21,000 with normalization to
the amount of protein in each sample. Because equimolar
quantities of Ȯ2
2 and cytochrome c react with one another, the
calculated values were taken as representative of Ȯ2
2 produced (13). The lower limit of detection of superoxide when
this technique is used is 0.75 mmol/ml (50 mmol/culture
plate).
H2O2 detection. H2O2 production was determined by two
approaches. In cell culture, H2O2 production was assessed
according to the method of Pick and Mizel (18) by use of an
automated enzyme immunoassay reader, with the following
modifications. BHMEC were incubated for 2 h (37°C) in the
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
Human plasma was obtained by centrifugation of blood
samples collected from normal fasting donors (both male and
female) at 1,500 g for 15 min at 4°C. Preliminary experiments
indicated no differences in response of isolated LDL from
male and female subjects to E2. All subjects were nonsmokers
and were not taking any antioxidants. Plasma lipoproteins
were sequentially isolated by ultracentrifugation, dialyzed,
and stored at 4°C after being purged with nitrogen as
described previously (23). Briefly, very low density lipoproteins, LDL, and HDL were isolated with solid potassium
bromide when density was ,1.006 at 42,000 rpm for 18 h,
when density was 1.006–1.063 at 42,000 rpm for 20 h, and
when density was 1.063–1.21 at 48,000 rpm for 24 h, respectively, with a Beckman L3-50 ultracentrifuge and a Ti 50.3
rotor. The lipoproteins were flushed with N2 and dialyzed
Dulbecco’s PBS (0.037 M Na2, HPO4, 7 H2O; 0.0018 M K2,
HPO4, 0.046 M NaCl, and KCl 0.003 M, pH 7.4) without
EDTA for 24 h with two changes of dialyzing solution. Protein
was determined by the method of Lowry et al. (12).
E1003
E1004
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
presence of 1.5 mM menadione, with or without E2, at varying
concentrations (0.1 µM-100 µM). In the second approach, we
used the chemical reaction between azobis and LDL. The
reaction mixture consisted of 0.6 ml of PRS buffer [140 mM
NaCl, 10 µM potassium phosphate buffer (pH 7.0), 5.5 mM
dextrose, 0.28 mM phenol red, and 8.5 U/ml (50 µg/ml)
horseradish peroxidase], 330 µg azobis, and 100 µg LDL,
which were incubated at 37°C for 0–2 h. The sample was then
placed in a cuvette with the addition of 10 µl of 1 M NaOH,
and absorption was read at 610 nm.
Hydroxyl radical detection. Effect of E2 on hydroxyl radical
formation was examined both in cell culture and by a
chemical method. In both cases the detection and quantification of hydroxyl radicals were performed by HPLC, as described by Onodera and Ashraf (17) with modifications. The
assay is based on the chemical interaction of ȮH with salicylic
acid, forming 2,3- and 2,4-DHBAs. The products can be
separated from one another and from salicylic acid by HPLC.
Both can be used as an index of ȮH production, which is
quantitated by integrating the area under the peak using
2,4-DHBA as an internal standard. The in vitro system of
generating hydroxyl radicals was examined by a modification
of the method of Onodera and Ashraf, as follows: 15 µM H2O2,
33 µM CuSO4, 10 µM salicylic acid, 10 µM 2,4-DHBA (internal standard), and 50 µl of 1 N HCl were placed in a 5-ml tube
with or without E2 and mixed for 120 s. E2 concentration
ranged from 10 to 100 µM. The sample was extracted with 4
ml of HPLC-grade diethyl ether on a Vortex mixer for 90 s.
The diethyl ether layer was separated and completely evaporated under nitrogen. The extraction was carried out a second
time. The dried residue was dissolved in 50 µl of 1 N HCl and
32.5 µl of mobile phase, and 20 µl of this solution were
injected into the HPLC unit. The amount of 2,5-DHBA was
calculated and expressed as nmol/mg protein. A Perkin-Elmer
Series 200 LC pump with Applied Biosystems 785A programmable ultraviolet (UV) detector at a wavelength of 315 nm
and a Hibar RT, Lichrosorb-RP-18 column (10 µm, 25 cm 3
0.4 cm) were used. The mobile phase consisted of 80% 0.03 M
citric acid-0.03 M acetic acid buffer (pH of 3.6) and 20%
methanol at a flow rate of 1.0 ml/min.
In cell culture, cells were plated in 60-mm wells and grown
to confluence. Cells were washed with PBS and placed in 3-ml
Eagle’s MEM (phenol red free, GIBCO/BRL) with 1% fetal
bovine serum. Salicylic acid (10 µl of 100 mM) and 200 µl
menadione (100 mM) were added for 1 h, and the samples
were incubated without and with E2 (0.01–1 mM) under UV
light for 30 min. The media were collected, and 200 µl of 1 N
HCl were added. 2,4-DHBA (50 µl of 1 mM) was used as an
internal standard. The sample was extracted twice with 5 ml
ethyl ether, dried down, and run on HPLC as described above.
Statistics
Data were expressed as means 6 SE. Group comparisons
were done by ANOVA with multiple comparison. A difference
of P , 0.05 was considered significant.
Table 1. Effect of selective oxygen species scavengers on
lipid peroxidation: role of E2
Incubation Condition
Maximum Absorption
% Inhibition
Control
Catalase
SOD
Mannitol
E2
E2 1 catalase
E2 1 SOD
1.191 6 0.01
0.914 6 0.01
0.729 6 0.01*
1.102 6 0.02
0.357 6 0.01*
0.457 6 0.02*
0.413 6 0.03*
060
23.50 6 0.67
39.50 6 0.50
7.50 6 0.50
70.50 6 0.50
61.50 6 0.51
65.50 6 0.67
Formation of conjugated dienes was measured by incubating 60 µg
low-density lipoprotein (LDL) with 5.5 mM/l AAPH in 0.985 ml PBS
medium. Absorbency at 234 nm was measured continuously in a
thermostat-controlled (37°C) spectophotometer. Increase in 234 nm
absorption was recorded every 5 min for 3 h. Values are means 6 SE
of duplicate determinations. Control, LDL incubated without any
antioxidants or scavengers; E2 , with estradiol-17b (54 µM); Catalase,
with catalase (0.1 mg/ml); SOD, with superoxide dismutase (0.1
mg/ml); Mannitol, with mannitol (5 mM). * P , 0.05 vs. control.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
Fig. 1. Time course of formation of
conjugated dienes during AAPH-induced oxidation of low-density lipoproteins (LDL). LDL (60 µg) were incubated with 5.5 mM/l AAPH in 0.985 ml
PBS medium. Absorbance at 234 nm
was measured continuously in a thermostat-controlled (37°C) spectrophotometer. Increase in 234 nm absorption
was recorded every 5 min for 3 h.
Selected oxygen scavengers and estradiol-17b (E2 ) were added, as described
in MATERIALS AND METHODS. Each data
point represents a mean of two determinations.
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
E1005
Table 3. Effect of scavengers of oxygen species on DNA
damage: role of E2
Incubation Conditions
% Damage
Control
E2 (36 µM)
E2 1 catalase 1 SOD
E2 1 catalase 1 mannitol
E2 1 catalase 1 SOD 1 mannitol
Catalase 1 mannitol 1 SOD
100.0 6 0
54.0 6 0.58
48.0 6 1.0
54.0 6 0.88
45.0 6 1.85
60.0 6 1.54
DNA damage was induced by H2O2 (15 µM) in combination with
Cu21 (33 µM), and protection afforded by E2 alone or in combination
with various scavengers was assessed as described in Fig. 2. Percent
damage is calculated as the total amount of open circular and linear
DNA formed. Values are means 6 SE of 3 determinations.
DNA Damage
Fig. 2. Agarose gel electophoresis (stained with ethidium bromide) of
OX-174 RFI plasmid DNA incubated with H2O2 and Cu21 in the
presence or absence of E2 at 37°C for 30 min. A, B, and C represent
supercoiled, open circular, and linear forms of DNA, respectively.
Lane 1: control DNA; lanes 2–4: DNA 1 H2O2 (15 µM) 1 Cu21 (33
µM); lanes 5, 6: DNA 1 H2O2 (15 µM) 1 Cu21 (33 µM) 1 E2 (36 µM).
RESULTS
Lipid Peroxidation
To examine the mechanism of E2 protection, we
determined the formation of conjugated dienes during
AAPH-induced oxidation of LDL (Fig. 1). Lag times for
E2 could not be estimated because of the nature of the
curve, as noted in our previous study (4). LDL peroxidation induced by azobis, a peroxy radical generator, was
significantly inhibited (39.5 6 0.71%) by SOD, a scavenger of superoxide (0.1 mg/ml). Incubation with catalase,
a hydrogen peroxide scavenger, at 0.1 mg/ml, resulted
in partial inhibition (23.5 6 2.1%) of LDL peroxidation.
Mannitol (5 mM), a hydroxyl radical scavenger, had no
Supercoiled plasmid OX-174 RFI DNA was examined
as a substrate, because relaxation of the molecule from
the supercoiled to the open circular or linear forms is an
indicator of single-stranded or double-stranded breaks,
respectively. Cu21 and H2O2 alone did not induce damage to the DNA. The addition of Cu21 with H2O2
induced a significant formation of linear DNA, form C
(Fig. 2). With the independent addition of catalase,
mannitol, and SOD, there was 49, 61, and 65% DNA
damage, respectively, compared with control (100%)
(Table 2). When all three free radical scavengers were
added together, 60% damage still occurred. From the
Table 2. Effect of inhibitors of oxygen species
on DNA damage
Incubation Conditions
% Damage
Control
Catalase (0.1 mg/ml)
Mannitol (5 mM)
SOD (0.1 mg/ml)
Sodium azide (5 mM)
Catalase 1 mannitol 1 SOD
100.0 6 0
49.0 6 1.77
61.0 6 4.67
65.0 6 3.53
100.0 6 0
60.0 6 1.54
DNA damage was induced by H2O2 (15 µM) in combination with
Cu21 (33 µM), and effect of addition of scavengers alone or in
combination on DNA damage was measured as described in legend to
Fig. 2. Percent damage is calculated as the total amount of open
circular and linear DNA formed. Values are means 6 SE of 3
determinations.
Fig. 3. Effect of E2 on O2
2 production by bovine heart microvessel
endothelial cells (BHMEC). Superoxide dismutase (SOD) production
was measured after incubation with BHMEC with Krebs-Ringer
bicarbonate (KRB) solution containing 20 µM ferricytochrome c.
Pairs of culture dishes with or without SOD were exposed to
experimental conditions at 37°C. Supernatants were collected and
absorbency was determined at 550 nm. Values are means 6 SD of
triplicate determinations.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
effect. The quantitative aspects of changes in lipid
peroxidation induced by the scavengers of ROS and
their comparison with E2-induced changes are shown in
Table 1. LDL peroxidation was markedly inhibited
(70.5 6 0.71%) by E2 (54 µM), which was greater than
that of any of the free radical scavengers examined
(Table 1). The combination of catalase or SOD with E2
did not result in any further degree of inhibition.
E1006
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
data (Table 2), it appears that hydrogen peroxide,
superoxide, and hydroxyl radicals are all contributing
to DNA damage. Sodium azide was completely ineffective in inhibiting damage, suggesting that singlet oxygen appears to play no role in the generation of DNA
damage. E2 significantly reduced the formation of linear DNA (Table 3). E2 alone was just as effective in
inhibiting DNA damage as the other individual scavengers alone or in combination. When E2 was added in
combination with the radical scavengers, no additional
protection was noted (Table 3).
Effect of E2 on Formation of ROS
The effect of E2 on ROS production (Ȯ2
2 , H2O2, and
ȮH) is shown in Figs. 3–7. In terms of superoxide
production in cell culture, E2 at 1 µM concentration was
37% effective in inhibiting Ȯ2
2 formation, with a doseresponse effect with increasing E2 concentrations (Fig.
3). E2 at concentrations up to 1 µM had no effect on the
production of H2O2 in cell culture, but a slight inhibition was noted at higher concentrations of E2 (Fig. 4). In
in vitro experiments using LDL, however, we found
marked inhibition of H2O2 production at 54 µM concen-
Fig. 5. Effect of E2 on H2O2 formation. LDL (100 µg) were incubated
at 37°C in 0.6 ml of PRS buffer with 330 µg azobis for 0–2 h with or
without E2 (54 µM). Samples were then placed in a cuvette with
addition of 10 µl (1 M NaOH), and absorption was read at 610 nm.
Values are means 6 SD of triplicate determinations.
Fig. 6. Effect of E2 on ȮH formation in BHMEC. Cells were washed
with PBS and placed in 3 ml MEM (phenol red free) (GIBCO/BRL)
with 1% fetal bovine serum. Salicylic acid (10 µl of 100 mM) and 200
µl menadione (100 mM) were added for 1 h, and samples were
incubated with and without E2 under ultraviolet (UV) light for 30
min. Medium was extracted with 5 ml ethyl ether after addition of
200 µl of 1 N HCl and was quantitated by HPLC using 2,4dihydroxybenzoic acid (DHBA; 50 µl of 1 mM) as a standard. DHBA
peak area/internal standard represents total amount of 2,3- and
2,5-DHBA formed in relation to a known amount of 2,4-DHBA
(internal standard).
tration of E2 (Fig. 5). E2 concentrations up to 100 µM
had no effect on production of ȮH in cell culture (Fig. 6),
but inhibition at a concentration of 10 µM E2 could be
seen when the chemical technique was utilized (Fig. 7).
DISCUSSION
In our present studies, we have examined the antioxidant properties of E2 by utilizing techniques that assess
lipid peroxidation and DNA damage. On the basis of
our studies examining the mechanism of inhibition of
lipid peroxidation and DNA damage by E2, three important points may be made. 1) It appears that lipid
peroxidation is most often induced by Ȯ2
2 and H2O2, as
Fig. 7. Effect of E2 on ȮH radical formation. H2O2 (15 µM), CuSO4 (33
µM), salicylic acid (10 µM), 2,4-DHBA (internal standard, 10 µM),
and 1 N HCl (50 µl) were placed in a 5-ml tube with or without E2 and
mixed for 120 s. Samples were extracted with diethyl ether and
evaporated under nitrogen. Residue was dissolved in 50 µl of 1 N HCl
and 32.5 µl of mobile phase, and 20 µl of this solution were injected
into HPLC as described in MATERIALS AND METHODS. 2,3- and 2,5DHBA represent salicylate derivatives formed. DHBA peak area/
internal standard represents total amount of 2,3 and 2,5-DHBA
formed in relation to a known amount of 2,4-DHBA (internal standard).
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
Fig. 4. Effect of E2 on H2O2 formation in BHMEC. BHMEC were
incubated for 2 h at 37°C in the presence of 1.5 mM menadione with
or without E2, and H2O2 was measured as described in MATERIALS AND
METHODS. Values are means 6 SD of triplicate determinations.
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
age either by inhibiting Ȯ2
2 production and/or by acting
as a chain-breaking antioxidant. The concentration of
E2 in our studies is slightly higher than the plasma
levels reached after estrogen therapy (0.2–1 3 1026
mol/l). We think that inhibition of lipid peroxidation
and DNA damage by E2 contributes to the cardioprotection noted after estrogen therapy. Furthermore, in view
of the potential role of superoxide in lipid peroxidation,
attempts to manipulate SOD in vivo would be beneficial.
This study was supported in part by National Heart, Lung and
Blood Institute Grant HL-50881.
Address for reprint requests: M. T. Ravi Subbiah, Univ. of Cincinnati, PO Box 670540, Cincinnati, OH 45267-0540.
Received 18 November 1997; accepted in final form 26 February
1998.
REFERENCES
1. Alvarez, R. J., Jr., S. J. Gips, N. Moldovan, C. C. Wilhide,
E. E. Milliken, A. T. Hoang, R. H. Hruban, H. S. Silverman,
C. V. Dang, and P. J. Goldschmidt-Clermont. 17b-Estradiol
inhibits apoptosis of endothelial cells. Biochem. Biophys. Res.
Commun. 237: 372–381, 1997.
2. Ames, B. N. Endogenous oxidative DNA damage, aging and
cancer. Free Radical Res. Commun. 7: 121–128, 1989.
3. Ames, B. N., and L. S. Gold. Endogenous mutagens and the
causes of aging and cancer. Mutation Res. 250: 3–16, 1991.
4. Ayres, S. A., M. Tang, and M. T. R. Subbiah. Estradiol-17b as
an antioxidant: some distinct features when compared with
common fat-soluble antioxidants. J. Lab. Clin. Med. 128: 367–
375, 1996.
5. Bhat, R., and S. M. Hadi. DNA damage by tannic acid and
Cu(II): sequence specificity of the reaction and involvement of
oxygen species. Mutation Res. 313: 39–48, 1994.
6. Brann, D. W., L. B. Hendry, and V. B. Mahesh. Emerging
diversities in the mechanism of action of steroid hormones. J.
Steroid Biochem. Mol. Biol. 52: 113–133, 1995.
7. Bush, T. L., E. Barrett-Connor, L. D. Cowan, M. H. Criqui,
R. B. Wallace, C. M. Suchindran, H. A. Tyroler, and B. M.
Rifkind. Cardiovascular mortality and non-contraceptive estrogen use in women: results from the Lipid Research Clinics
Program Follow-up Study. Circulation 75: 1002–1009, 1987.
8. Esterbauer, H., G. Striegl, and H. Puhl. Continuous monitoring of in vitro oxidation of human low-density lipoprotein. Free
Radical Res. Commun. 6: 67–75, 1989.
9. Keaney, J. F., Jr., G. T. Shwaery, A. Xu, R. J. Nicolosi, J.
Loscalzo, T. L. Foxall, and J. A. Vita. 17-b Estradiol preserves
endothelial vasodilator function and limits low-density lipoprotein oxidation in hypercholesterolemic swine. Circulation 89:
2251–2259, 1994.
10. Lacort, M., A. M. Leal, M. Liza, C. Martin, R. Martinez, and
M. B. Ruiz-Larrea. Protective effect of estrogens and catecholestrogens against peroxidative membrane damage in vitro. Lipids
30: 141–146, 1995.
11. Li, Y., and M. A. Trush. DNA damage resulting from the
oxidation of hydroquinone by copper: role for a Cu(II)/Cu(I) redox
cycle and reactive oxygen generation. Carcinogenesis 14: 1303–
1311, 1993.
12. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J.
Randall. Protein measurement with the Folin phenol reagent. J.
Biol. Chem. 193: 265–275, 1951.
13. Matsubara, T., and M. Ziff. Superoxide anion release by
human endothelial cells: synergism between a phorbol ester and
a calcium ionophore. J. Cell. Physiol. 127: 207–210, 1986.
14. Maziere, C., M. Auclair, M. C. Ronveaux, S. Salmon, R.
Santus, and J. C. Maziere. Estrogens inhibit copper and
cell-mediated modification of low density lipoprotein. Atherosclerosis 89: 175–182, 1991.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
indicated by our studies showing that 60% of the
damage is inhibited by SOD and catalase. The remaining amount of damage appears to be caused by peroxyl
radicals, which are the initial radicals produced by the
oxidation initiator azobis (16). E2 alone collectively
blocks 70% of the damage, and the further addition of
the two scavengers provides no additional inhibition of
damage. This indicates that E2 is acting as a chainbreaking antioxidant, inhibiting the effect of H2O2, Ȯ2
2,
and hydroperoxyl radicals. 2) E2 action in inhibiting
DNA damage supports this view. E2 prevented DNA
strand breaks in a manner similar to the free radical
scavengers catalase, SOD, and mannitol. When all
three scavengers were added together and in combination with E2, the formation of DNA strand breaks could
not be entirely prevented. Thus the action of E2 does not
selectively restrict itself to specific ROS but prevents
the continuation of the propagation phase of free
radical attack on its intended target. Because sodium
azide had no effect on inhibiting damage, the singlet
oxygen did not account for any of this disparity. 3) The
experiments concerned with the effect of E2 on the
formation of ROS demonstrate that E2 has significant
effect on the formation of Ȯ2
2 in cell culture. This
observation differs from that of Rifici and Khachadurian (19), who could not demonstrate E2’s effect on Ȯ2
2
production in utilizing mononuclear cells. However,
using BHMEC, we could demonstrate that increasing
concentrations of E2 were effective in blocking Ȯ2
2
production. This difference could be attributed to the
utilization of different kinds of cells. In cell culture we
do not see any significant effect of E2 on the formation
of either H2O2 or ȮH radicals. In the chemical method,
high E2 concentrations (50 µM) seem to have an effect.
At lower concentrations, E2 appears to have no effect on
H2O2 and ȮH formation.
Our in vitro studies suggest that E2 might be preventing oxidative DNA damage to some extent by inhibiting
the formation of superoxides. The in vivo significance of
this finding deserves some discussion in view of a
recent report stating that E2 decreases apoptosis of
endothelial cells (1). In cellular apoptosis the BCL-2
gene plays a central role, and a variety of stimuli such
as oxidants, toxins, oncogenes, and some growth factors
can modulate expression of this gene (25). Estrogens
are known to modulate the transcription of a number of
genes through their binding to cytosolic estrogen receptors, which translocate to nucleus. The receptor/
estrogen complex binds to specific palidromic DNA
targets (6). It is possible that, in this way, estrogens can
directly or indirectly modulate BCL-2 expression. In
amyotrophic lateral sclerosis, cell death is considered
to be due to a mutation in SOD, causing inability to
handle oxygen radicals (20). In vitro superoxide-related
cell death can be corrected by antioxidants (26). Therefore, it is possible that the ability of estrogens to
decrease oxidation of DNA damage in vitro might have
some in vivo significance in terms of apoptosis.
On the basis of our studies, we conclude that E2
might be decreasing lipid peroxidation and DNA dam-
E1007
E1008
ESTROGEN, LIPID PEROXIDATION, AND DNA DAMAGE
21. Sack, M. N., D. J. Rader, and R. O. Cannon III. Oestrogen and
inhibition of low-density lipoproteins in postmenopausal women.
Lancet 343: 269–270, 1994.
22. Stampfer, M. J., G. A. Colditz, W. C. Willet, J. E. Manson, B.
Rosner, F. E. Speizer, and C. H. Hennekens. Postmenopausal
estrogen therapy and cardiovascular disease. N. Engl. J. Med.
325: 756–762, 1991.
23. Subbiah, M. T. R., B. Kessel, M. Agrawal, R. Rajan, W.
Abplanalp, and Z. Rymazewski. Antioxidant potential of
specific estrogens on lipid peroxidation. J. Clin. Endocrinol.
Metab. 77: 1095–1098, 1993.
24. Tang, M., and M. T. R. Subbiah. Estrogens protect against
hydrogen peroxide and arachidonic acid induced DNA damage.
Biochim. Biophys. Acta 1299: 155–159, 1996.
25. Thompson, C. B. Apoptosis in the pathogenesis and treatment
of disease. Science 267: 1456–1462, 1995.
26. Troy, C. M., and M. L. Shelanski. Down-regulation of copper/
zinc superoxide dismutase causes apoptotic death in PC12
neuronal cells. Proc. Natl. Acad. Sci. 91: 6384–6387, 1994.
27. Vaca, C. E., J. M. Wilhelm, and M. Harms-Kingdahl. Interaction of lipid peroxidation byproducts with DNA, a review. Mutat.
Res. 195: 137–144, 1988.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.4 on June 18, 2017
15. McCord, J. M., and I. Fridovich. Superoxide dismutase: an
enzymatic function for erythrocuprein (hemocuprein). J. Biol.
Chem. 244: 6049–6063, 1969.
16. Niki, E., M. Saito, Y. Yoshikawa, Y. Yamamoto, and Y.
Kamiya. 2,28-Azobis[2-amidinopropane]dihydrochloride is a water soluble peroxy-radical generator. Bull. Chem. Soc. Jpn. 59:
471–477, 1986.
17. Onodera, T., and M. Ashraf. Detection of hydroxyl radicals in
the post-ischemic reperfused heart using salicylate as a trapping
agent. J. Mol. Cell. Cardiol. 23: 365–370, 1991.
18. Pick, E., and D. Mizel. Rapid microassays for the measurement
of superoxide and hydrogen peroxide production by macrophages
in culture using an automatic enzyme immunoassay reader. J.
Immunol. Methods 46: 211–226, 1981.
19. Rifici, V. A., and A. K. Khachadurian. The inhibition of
low-density lipoprotein oxidation by 17-b estradiol. Metabolism
41: 1110–1114, 1992.
20. Rosen, D. R., T. Siddique, D. Patterson, D. A. Figlewicz, P.
Sapp, A. Hentati, D. Donaldson, J. Goto, J. P. O’Regan, H.
Deng, et al. Mutations in Cu/Zn superoxide dismutase gene are
associated with familial amyotrophic lateral sclerosis. Nature
362: 59–62, 1993.