Physiological Testosterone Replacement Therapy Attenuates
Fatty Streak Formation and Improves High-Density
Lipoprotein Cholesterol in the Tfm Mouse
An Effect That Is Independent of the Classic Androgen Receptor
Joanne E. Nettleship, BSc, PhD; T. Hugh Jones, BSc, MBChB, MD, FRCP;
Kevin S. Channer, BSc, MBChB, MD, FRCP; Richard D. Jones, BSc, PhD
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Background—Research supports a beneficial effect of physiological testosterone on cardiovascular disease. The
mechanisms by which testosterone produces these effects have yet to be elucidated. The testicular feminized (Tfm)
mouse exhibits a nonfunctional androgen receptor and low circulating testosterone concentrations. We used the Tfm
mouse to determine whether testosterone modulates atheroma formation via its classic signaling pathway involving the
nuclear androgen receptor, conversion to 17␤-estradiol, or an alternative signaling pathway.
Methods and Results—Tfm mice (n⫽31) and XY littermates (n⫽8) were separated into 5 experimental groups. Each group
received saline (Tfm, n⫽8; XY littermates, n⫽8), physiological testosterone alone (Tfm, n⫽8), physiological
testosterone in conjunction with the estrogen receptor ␣ antagonist fulvestrant (Tfm, n⫽8), or physiological testosterone
in conjunction with the aromatase inhibitor anastrazole (Tfm, n⫽7). All groups were fed a cholesterol-enriched diet for
28 weeks. Serial sections from the aortic root were examined for fatty streak formation. Blood was collected for
measurement of total cholesterol, high-density lipoprotein cholesterol (HDLC), non-HDLC, testosterone, and 17␤estradiol. Physiological testosterone replacement significantly reduced fatty streak formation in Tfm mice compared
with placebo-treated controls (0.37⫾0.07% versus 2.86⫾0.39%, respectively; Pⱕ0.0001). HDLC concentrations also
were significantly raised in Tfm mice receiving physiological testosterone replacement compared with those receiving
placebo (2.81⫾0.30 versus 2.08⫾0.09 mmol/L, respectively; P⫽0.05). Cotreatment with either fulvestrant or
anastrazole completely abolished the improvement in HDLC.
Conclusion—Physiological testosterone replacement inhibited fatty streak formation in the Tfm mouse, an effect that was
independent of the androgen receptor. The observed increase in HDLC is consistent with conversion to 17␤-estradiol.
(Circulation. 2007;116:2427-2434.)
Key Words: atherosclerosis 䡲 cholesterol 䡲 hormones 䡲 lesion
T
raditionally, testosterone has been considered detrimental to the cardiovascular system because of reports of
sudden death from the abuse of high-dose anabolic steroids1
and the higher male incidence of coronary artery disease.2
Current research suggests, however, that low testosterone
rather than testosterone per se is associated with coronary
artery disease and that a relative deficiency in these circulating levels may lead to a proatherosclerotic environment
(reviewed elsewhere3).
terol accumulation in cholesterol-fed, orchidectomized male
rabbits4 – 6 and low-density lipoprotein (LDL) receptor knockout mice.7 Similarly, treatment with the male androgen
dihydroepiandrosterone also has been reported to attenuate
atheroma formation in castrated cholesterol-fed rabbits.6,8 –10
Testosterone is reported to inhibit neointimal growth in
endothelial-denuded rabbit aortic rings.11 Contrary to these
studies, 1 study demonstrated a deleterious action in male
mice. Von Dehn et al12 reported a decrease in atheroma
formation in the aortic sinus and ascending aorta after
testosterone suppression by administration of the GnRH
antagonist Cetrorelix, whereas testosterone treatment led to a
small but significant increase in lesion size. The reason for
this discrepancy is unclear; however, the majority of these
Clinical Perspective p 2434
Beneficial effects of androgen supplementation in atherogenesis have been described in 4 animal studies. Testosterone
replacement has been demonstrated to prevent aortic choles-
Received April 19, 2007; accepted August 7, 2007.
From the Academic Unit of Diabetes, Endocrinology and Metabolism, University of Sheffield, Sheffield, UK (J.E.N., T.H.J., R.D.J.); Centre for
Diabetes and Endocrinology, Barnsley Hospital NHS Foundation Trust, Barnsley, UK (T.H.J.); and Department of Cardiology, Royal Hallamshire
Hospital, and Biomedical Research Centre, Sheffield Hallam University, Sheffield, UK (K.S.C.).
Correspondence to Dr Joanne E Nettleship, Academic Unit of Diabetes, Endocrinology & Metabolism, University of Sheffield, Sheffield, UK. E-mail
[email protected]
© 2007 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org
DOI: 10.1161/CIRCULATIONAHA.107.708768
2427
2428
Circulation
November 20, 2007
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
studies provide evidence for a protective role for testosterone
against atherogenesis in male animal models, an action that is
independent of changes in serum lipid levels. It has been
speculated that the antiatherogenic effects of testosterone are
mediated exclusively by its conversion to 17␤-estradiol,
although within the literature only 1 article directly supports
this hypothesis.7
Growing evidence supports the ability of sex hormones to
elicit nongenomic effects (ie, responses not mediated through
the classic nuclear androgen receptor [AR] but instead initiated by hypothetical androgen-specific membrane receptors).
These nongenomic actions are characterized mainly by rapid
onset and a lack of sensitivity toward inhibitors of transcription and protein synthesis (reviewed elsewhere13). For example, testosterone has been demonstrated to elicit immediate
vasodilatation in a number of vascular beds through blockade
of an L-type calcium channel, a mechanism that is independent of the nuclear AR.14 Furthermore, the ability of testosterone to regulate cellular activities independently of the
classic genomic route is not confined to the vascular smooth
muscle cell. Many studies have revealed the modulation of
calcium channel activity by steroids in the membrane of
several different cell systems.15–18 However, the involvement
of these nongenomic actions in the observed antiatherosclerotic effects of testosterone is currently unknown.
The aims of the present study were to determine whether
testosterone deficiency and/or the absence of a functional AR
is associated with increased atherosclerosis after cholesterol
feeding, to determine whether physiological testosterone
replacement modulates atherosclerosis by AR-dependent or
AR-independent processes, and to clarify the involvement of
17␤-estradiol in any action of testosterone.
Methods
The Tfm Mouse
In the present study, we used the Tfm mouse, which exhibits an
X-linked, single– base-pair deletion in the gene encoding the classic
AR.19,20 This deletion causes a frameshift mutation in the AR
mRNA, resulting in a stop codon in the amino-terminal region,
leading to premature termination of AR protein synthesis. On
translation, a truncated receptor protein is produced that lacks both
DNA- and steroid-binding domains.19,20 Unlike the clinical situation
of androgen insensitivity in men, serum levels of testosterone are
reduced in the Tfm mouse compared with controls, reportedly
10-fold lower in Tfm mice compared with XY littermates (⬇1.5
compared with 15 nmol/L, respectively).21 The lack of androgen
production by the testes in Tfm mice is due primarily to loss of
17␣-hydroxylase, an enzyme conduit to the steroidogenesis
pathway.22
Protocol 1: Determination of the Influence of AR
Dysfunction and Testosterone Deficiency on
Atheroma Formation
Preliminary studies have demonstrated that Tfm mice develop aortic
fatty streak formation and a proatherogenic lipid profile after being
fed a cholesterol-enriched diet ad libitum for a 28-week period, an
effect that was not observed in XY littermate controls.23 However,
because of the combined AR dysfunction and the testosterone
deficiency inherent in these animals, these preliminary data do not
allow determination of which of these 2 abnormalities is responsible
for the observed effect. To address this deficit, the following
experiments were undertaken.
At 8 weeks of age, the 2 groups of mice, Tfm mice (n⫽8) and XY
littermates (n⫽9), underwent either sham operation or surgical
castration, respectively. Animals were allowed to recover in individual cages for 2 weeks; at 10 weeks of age, they were fed a
cholesterol-enriched diet ad libitum (42% butterfat, 1.25% cholesterol, and 0.5% cholate; Special Diet Services, Essex, UK) for a
period of 28 weeks. In addition, at 10 weeks of age, another 2 groups
of nonsurgically prepared Tfm mice (n⫽4) and XY littermates (n⫽4)
were fed a cholesterol-enriched diet ad libitum (42% butterfat, 1.25%
cholesterol, and 0.5% cholate; Special Diet Services, UK) for 28
weeks. All animals were carefully monitored for the duration of the
study and weighed on a weekly basis.
Protocol 2: Determination of a Dosing Regimen to
Replace Testosterone to Physiological Levels
in the Tfm Mouse
To assess the impact of testosterone replacement therapy on fatty
streak formation in the Tfm mouse, we first had to establish and
maintain a dosing regimen to replace testosterone to physiological
concentrations. Experiments were undertaken in nonsurgically prepared mice to determine the appropriate volume and frequency of
100 mg/mL testosterone esters (Sustanon100; testosterone propionate 20 mg/mL, testosterone phenylpropionate 40 mg/mL, and
testosterone isocaproate 40 mg/mL, Organon Laboratories Ltd,
Cambridge, UK) required to replace testosterone to physiological
concentrations. The human physiological replacement dose of Sustanon100 is 3.5 mg/kg administered via a once-fortnightly intramuscular injection. Because of the increased metabolic rate of the
smaller species of mammal, an adjustment of the human dose by
between 5 and 15 times is commonly used. Consequently, 8-weekold Tfm mice received a single 10-␮L intramuscular injection of 100
mg/mL testosterone (Sustanon100), providing a dose of 50 mg/kg
(⬇14 times the human dose). Mice were killed at serial intervals at
1, 2, 4, 7, 10, or 14 days after injection via a Home Office–approved
schedule 1 method, and the testosterone concentration was measured.
To ensure reproducibility of this pharmacokinetic profile, 18 additional Tfm mice received a second injection at day 14 and were then
killed on day 15, 16, 18, or 21.
Protocol 3: Determination of the Effect of
Physiological Testosterone Replacement and
the Influence of the AR and 17␤-Estradiol
on Atheroma Formation
At 9 weeks of age, Tfm mice (n⫽32) and XY littermates (n⫽8) were
randomly separated into 5 experimental groups. Each group received
a once-fortnightly intramuscular injection of 10 ␮L saline (Tfm,
n⫽8; XY littermates, n⫽8), 10 ␮L of 100 mg/mL testosteroneSustanon100 (Tfm, n⫽8), 10 ␮L of 100 mg/mL testosteroneSustanon100 in conjunction with 30 ␮L of 50 mg/mL of the estrogen
receptor ␣ (ER␣) antagonist fulvestrant (Faslodex, AstraZeneca,
Cheshire, UK) at 15 times the human dose (Tfm n⫽8), and 10 ␮L of
100 mg/mL testosterone-Sustanon100 in conjunction with 10 mg ·
kg⫺1 · d⫺1 of the aromatase inhibitor anastrazole (Arimidex, AstraZeneca) at 10 times the human dose in drinking water (Tfm, n⫽7).
At 10 weeks of age, all groups were fed a cholesterol-enriched diet
ad libitum (42% butterfat, 1.25% cholesterol, and 0.5% cholate,
Special Diet Services, UK) for a period of 28 weeks.
Quantification of Fatty Streak Formation
At the end of the feeding period, all animals were killed via a Home
Office–approved schedule 1 method. The heart with thoracic aorta
attached was carefully dissected free from the adventitia. The basal
half of the ventricles and the ascending aorta were removed, and the
heart was perfused with physiological saline solution to remove any
blood clots. The upper half of the heart with the aortic root attached
was embedded in optical cutting temperature compound (Agar
Scientific Limited, Essex, UK) and frozen at ⫺80°C until processing
for analysis of lipid accumulation in the aortic sinus. Subsequently,
serial 8-␮m transverse frozen sections of the aortic root with valves
Nettleship et al
A
D
B
E
Testosterone Therapy and Fatty Streak Formation
2429
C
F
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
(50 per mouse) were cut on a cryostat, and 5 (every 10th section)
sections of the aortic root were used per animal. The frozen 8-␮m
cryosections of the aortic root were allowed to air dry onto charged
slides (Surgipath Europe, Peterborough, UK) at room temperature.
The sections were then rinsed with 60% isopropanol and stained for
15 minutes with Oil Red O. Slides were then rinsed in 60%
isopropanol until the slides appeared colorless, washed in distilled
water, and counterstained with hematoxylin for 1 minute. The slides
were then washed well in tap water, rinsed in distilled water, and
mounted with glycerol gelatin (Sigma, Poole, Dorset, UK). After Oil
Red O staining, aortic root sections were digitally photographed.
Quantification of the lipid-stained areas was performed with
computer-assisted morphometry using Scion Image software (NIHScion Image, National Institutes of Health, Bethesda, Md). Lipidstained areas were expressed as percentage of the intimal plus medial
area, with an average value calculated for each of the 5 sections
analyzed per animal.
Blood Collection
At the end of the relevant feeding or dosing period, all animals were
killed via a Home Office–approved schedule 1 method. Following
midline sternotomy, the rib cage was opened and whole blood was
collected from the chest cavity after severance of the thoracic aorta.
Whole blood for serum measurements was drawn through a 1-mL
syringe (Becton Dickinson, Franklin Lakes, NJ), collected into
1.5-mL Eppendorf tubes, and centrifuged at 3000 rpm for 10
minutes; the serum was frozen at ⫺80°C. The following analyses
were made on individual serum samples.
Measurement of Total Testosterone and 17␤-Estradiol
Serum quantification of total testosterone (DRG Instruments GmBH,
Marburg, Germany) and 17␤-estradiol (Oxford Biomedical Research, Oxford, Mich) was measured in duplicate via ELISA.
Measurement of Total and High-Density
Lipoprotein Cholesterol
Serum quantification of total cholesterol (TC) and high-density
lipoprotein cholesterol (HDLC) was measured in duplicate in a
blinded fashion with an automated colorimetric assay by the
Department of Clinical Chemistry, Royal Hallamshire Hospital,
Sheffield, UK.
Calculated Measurements
Non-HDLC was calculated by subtracting HDLC from TC.
Statistical Analysis
In accordance with internal and external ethical review policies
(University of Sheffield and UK Home Office, respectively), the
number of animals used in the study was determined in the
Figure 1. Typical photographs showing lipid
deposition within the aortic root stained with
Oil Red O and counterstained with hematoxylin
in a Tfm mouse (A and B) vs XY littermates (C),
sham-operated Tfm mouse (D), and castrated
XY male (E) after feeding on a high-cholesterol
diet alone for 28 weeks and in Tfm mice receiving physiological testosterone replacement therapy in conjunction with feeding on a high-cholesterol diet for 28 weeks (F).
preliminary studies and reflects this minimum level of animal usage
without compromising the statistical validity of the data.
Results are expressed as mean⫾SEM. All data were demonstrated
to be normally distributed as assessed by the Kolmogorov-Smirnov
test for normality, which was confirmed using probability plots
contained within the SPSS package (SPSS Inc, Chicago, Ill). Data
were subsequently analyzed by 1-way ANOVA, with subsequent
analysis made with Student paired or unpaired t test, as appropriate,
to identity individual differences between groups. When multiple t
tests were applied, to allow for type 1 errors, the probability value
was modified by use of the Bonferroni correction, and the probability
values displayed in the text are those obtained after the use of this
correction. Significance was assumed with values of P⬍0.05. All
procedures were carried out under the jurisdiction of the UK Home
Office project license (project license number, 40/2227; personal
license number, 40/7385), governed by the UK Animals Scientific
Procedures Act 1986.
The authors had full access to and take full responsibility for the
integrity of the data. All authors have read and agree to the
manuscript as written.
Results
Protocol 1: Determination of the Influence of
AR Dysfunction and Testosterone Deficiency
on Atheroma Formation
Marked fatty streak formation was observed within the aortic
root of the sham-operated Tfm mice, the castrated XY
littermates, and the nonsurgically prepared Tfm mice but not
the nonsurgically prepared XY littermates (Figures 1 and 2).
The amount of lipid deposition in the castrated XY males,
sham-operated Tfm, and nonsurgically prepared Tfm mice
(4.39⫾0.26%, 2.15⫾0.28%, and 3.74⫾0.72%, respectively)
was similar although marginally significantly different in that
the sham-operated Tfm mice had less lipid deposition than
the nonsurgically prepared Tfm animals (Figures 1 and 2).
The amount of lipid deposition in the castrated XY littermates, however, was markedly greater than that observed in
nonsurgically prepared XY littermates fed the same diet for
the same amount of time (4.39⫾0.26% versus 0.31⫾0.11%
respectively; P⬍0.0001) (Figures 1 and 2).
Circulating serum levels of TC, HDLC, and non-HDLC
were statistically similar in sham-operated Tfm mice compared with the nonsurgically prepared Tfm mice. However,
2430
Circulation
November 20, 2007
6.00
*
5.50
Fatty Streak Formation (%)
5.00
4.50
Figure 2. Bar chart showing lipid deposition within
the aortic root (expressed as a percentage of the
medial area of 5 sections per animal) in nonsurgically prepared Tfm mice (n⫽4), sham-operated Tfm
mice (n⫽6), nonsurgically prepared XY littermates
(n⫽4), and castrated XY littermates (n⫽8) after
feeding on a cholesterol-enriched diet for 28
weeks. *P⬍0.05 via ANOVA.
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
Tfm
1
Sham-operated
Tfm
Diet alone
XY
Castrated
XY
Diet alone
A significant reduction in fatty streak formation was observed
in Tfm mice after testosterone replacement therapy compared
with placebo-treated Tfm mice (0.37⫾0.4% versus
2.86⫾0.39%, respectively; P⬍0.0001) (Figures 1 and 4).
After testosterone treatment, the level of lipid deposition
within the aortic root of the Tfm mouse was almost identical
to that observed in the placebo-treated XY littermates
(0.37⫾0.4% versus 0.32⫾0.06%, respectively; P⫽0.9). The
degree of lipid deposition in the Tfm mice treated with
physiological testosterone replacement therapy was significantly higher after cotreatment with either the ER␣ antagonist
fulvestrant or the aromatase inhibitor anastrazole compared
with Tfm mice treated with physiological testosterone replacement therapy alone (1.04⫾0.32% and 0.88⫾0.18%
versus 0.37⫾0.4%; both P⬍0.05). However, both were still
significantly lower than that of placebo-treated Tfm mice
(1.04⫾0.32% and 0.88⫾0.18% versus 2.86⫾0.39%; both
P⬍0.0001) (Figures 1 and 4).
No significant differences in serum levels of TC and
non-HDLC were observed between Tfm mice receiving
physiological testosterone replacement therapy or placebo.
However, a significant increase in serum levels of HDLC was
observed after testosterone replacement therapy in Tfm mice
Protocol 2: Determination of a Dosing Regimen to
Replace Testosterone to Physiological Levels
in the Tfm Mouse
Serum testosterone levels rose significantly from baseline in
Tfm mice after the 10-␮L intramuscular injection of 100
mg/mL testosterone at day 1 (Figure 3) and remained significantly higher throughout the remainder of the 14-day period.
Testosterone concentrations were observed to be above the
subphysiological range for most of the 14-day period. The
mean area under the curve for this 14-day period was 19.4
nmol/L. A similar window of activity was observed after the
second 10-␮L intramuscular injection of 100 mg/mL testosterone at day 14, with statistically similar testosterone concentrations observed. The mean area under the curve for this
second 14-day period was 18.8 nmol/L. These data revealed
that a single once-fortnightly intramuscular injection of testosterone (100 mg/mL Sustanon100) was sufficient to produce consistent physiological testosterone concentrations in
the Tfm mouse.
Serum testosterone concentrati on (nM)
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Protocol 3: Determination of the Effect of
Physiological Testosterone Replacement and
the Influence of the AR and 17␤-Estradiol
on Atheroma Formation
serum levels of TC, HDLC, and non-HDLC were markedly
increased in the castrated XY males after feeding on a
cholesterol-enriched diet for 28 weeks. This increase was
significantly greater than that observed in the nonsurgically
prepared XY littermates (15.7⫾4.9 versus 4.2⫾0.5 mmol/L,
P⬍0.05; 3.5⫾0.8 versus 2.2⫾0.4 mmol/L, P⬍0.05; and
12.2⫾4.6 versus 2.1⫾0.2 mmol/L, P⬍0.05, respectively).
70.0
60.0
Figure 3. Line graph showing serum concentrations of total testosterone in Tfm mice at baseline
and after an initial 10-␮L intramuscular injection of
100 mg/mL testosterone at day 0 (n⫽31; unbroken
line). An additional 10-␮L intramuscular injection
of 100 mg/mL testosterone was administered at
day 14 (n⫽18; broken line). The line graph below
shows the first- and second-cycle pharmacokinetic curves overlaid to demonstrate “visually” the
reproducibility of these data. The area between the
hashed parallel lines denotes the physiological
range of testosterone (10 to 30 nmol/L).
50.0
40.0
30.0
20.0
10.0
0.0
0
2
4
6
Time (days) post injection
8
10
12
14
Nettleship et al
Testosterone Therapy and Fatty Streak Formation
2431
6.00
5.50
5.00
% Lipid deposition
4.50
*
4.00
3.50
3.00
*
2.50
2.00
†
1.50
†
1.00
0.50
0.00
Tfm
Placebo
XY
Placebo
Tfm
1
S100
Tfm S100 +
Fulvestrant
Tfm S100 +
Anastrazole
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
compared with placebo-treated Tfm mice (2.81⫾0.30 versus
2.08⫾0.09 mmol/L; P⬍0.05). Serum levels of HDLC in Tfm
mice receiving physiological testosterone replacement therapy were comparable to those observed in the placebo-treated
XY littermates (2.81⫾0.30 versus 3.13⫾0.20 mmol/L respectively, P⫽0.4).
Serum levels of HDLC were lower in Tfm mice receiving
physiological testosterone replacement therapy in conjunction with the aromatase inhibitor anastrazole or the ER␣
antagonist fulvestrant compared with Tfm mice receiving
physiological testosterone replacement therapy alone, although this failed to reach statistical significance. However,
when expressed more accurately as a percentage of TC,
serum levels of HDLC were similarly reduced in Tfm mice
receiving physiological testosterone replacement therapy in
conjunction with fulvestrant or anastrazole to the levels
previously reported in placebo-treated Tfm mice (Figure 5).
Discussion
This study has clearly demonstrated that low endogenous
testosterone, be it associated with a nonfunctional AR (Tfm
mouse) or a fully functional AR (castrated XY male), is
associated with both fatty streak formation within the aortic
root and a proatherogenic circulating lipid profile after
feeding on a cholesterol-enriched diet. These observations
demonstrate that fatty streak formation is a consequence of
low endogenous testosterone levels in these mice and not a
result of the absence of a functional AR because aortic fatty
streaks were observed in surgically castrated male animals.
After cholesterol feeding, similar increases in serum TC,
HDLC, and non-HDLC were observed in the sham-operated
Tfm mice compared with the nonsurgically prepared Tfm
mice, with the greatest proportion of cholesterol being found
within the non-HDLC fraction. Similarly, castrated male
mice exhibited significant increases in TC, HDLC, and
non-HDLC, although this was significantly higher than that
observed in both nonsurgically prepared males and shamoperated Tfm mice. However, despite these marked differences in the circulating lipid profile in the castrated male
littermates compared with other experimental groups, the
degree of fatty streak formation in these animals was not
markedly elevated, also suggesting that the absence of testosterone rather than an abnormal lipid profile was the
primary cause of the observed lipid deposition. However, the
fact that marked differences in the circulating lipid profile
were observed in the castrated male animals compared with
the sham-operated Tfm mice suggests that AR-dependent
processes closely regulate cholesterol metabolism.
These data provide evidence that low serum testosterone is
linked to increased fatty streak formation. These data are in
agreement with a number of animal studies that describe a
deleterious relationship between aortic atherosclerosis and
low endogenous testosterone levels in male animal models of
atherosclerosis. Alexandersen et al6 used 100 sexually mature
*
80.0
*
70.0
Serum HDLC as % of TC
Figure 4. Bar chart showing lipid deposition within
the aortic root (expressed as a percentage of the
medial area of 5 sections per animal) in placebotreated Tfm mice (n⫽8), placebo-treated XY littermates (n⫽8), Tfm mice receiving physiological testosterone replacement with Sustanon100 (S100;
n⫽8), Tfm mice receiving physiological testosterone replacement with S100 in conjunction with
fulvestrant (n⫽8), and Tfm mice receiving physiological testosterone replacement with S100 in conjunction with anastrazole (n⫽7) after feeding on a
cholesterol-enriched diet for 28 weeks. *P⬍0.05 vs
indicated group; †P⬍0.0001 vs placebo-treated
Tfm mice via Student unpaired t test.
Figure 5. Bar chart showing serum concentrations
of HDLC as a percentage of TC in placebo-treated
Tfm mice (n⫽8), Tfm mice receiving physiological
testosterone replacement with Sustanon100 (S100;
n⫽8), Tfm mice receiving physiological testosterone replacement with S100 in conjunction with
fulvestrant (n⫽8), and Tfm mice receiving physiological testosterone replacement with S100 in conjunction with anastrazole (n⫽7) after feeding on a
cholesterol-enriched diet for 28 weeks. *P⬍0.05 vs
indicated group via Student unpaired t test.
60.0
50.0
40.0
30.0
20.0
10.0
0.0
Tfm
placebo
Tfm
S100
1
Tfm S100
+ Fulvestrant
Tfm S100
+ Anastrazole
2432
Circulation
November 20, 2007
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
male rabbits to examine the effects of natural androgens on
lipids. One arm of the study involved bilateral castration of 20
male rabbits, whereas 20 were sham operated. Animals were
fed an atherogenic diet for a 30-week period to induce aortic
atherosclerosis. A doubling of aortic plaque formation was
reported after castration in male rabbits compared with
sham-operated controls.6 Similarly, Nathan et al7 assessed the
effects of testosterone on early atherogenesis in orchidectomized LDL receptor knockout mice fed a cholesterolenriched diet for a period of 8 weeks. In one arm of the study,
the extent of lesion formation in the aortic region of male
mice with testosterone levels in the subphysiological range
(orchidectomized) was reported to be exacerbated compared
with testes-intact animals.7 A number of clinical studies also
have examined the association between serum levels of
testosterone and coronary artery disease in men.24 –28 These
studies differ in their definition of atherosclerosis and in size
but are consistent in their conclusion that endogenous serum
levels of testosterone are not raised in men with coronary
artery disease. In fact, the majority of these studies report that
serum testosterone levels are reduced in men with coronary
artery disease. Furthermore, low endogenous testosterone in
men also is reported to be associated with high serum levels
of TC and/or LDL cholesterol,29 –33 suggesting an association
between hypotestosteronemia and a proatherogenic lipid profile, consistent with the observations of the present study.
Testosterone replacement therapy results in a decrease in TC
in men with coronary artery disease34 and diabetes mellitus.35
Another important finding of the present study was that
long-term physiological testosterone replacement therapy induced a significant reduction in fatty streak formation in the
Tfm mouse compared with placebo-treated controls. This
provides further supporting evidence for a protective role of
testosterone in the development of fatty streak formation. In
fact, this reduction was so marked that the level of lipid
deposition observed was statistically similar to that of the
wild-type, placebo-treated XY littermates. These data demonstrate an inhibitory effect of testosterone on fatty streak
formation by an action that is independent of the AR because
these animals do not express a functional receptor protein.
However, a component of the effect of testosterone was
directly attributable to 17␤-estradiol. Testosterone is readily
converted to 17␤-estradiol via the enzyme aromatase. After
conversion, 17␤-estradiol has the potential to elicit cellular
effects by 3 potential mechanisms: activation of ER␣, activation of ER␤, and activation of nongenomic pathways
independently of these 2 classic genomic receptor pathways.
In the present study, the degree of aortic lipid deposition
observed in Tfm mice receiving physiological testosterone
replacement therapy was significantly increased by cotreatment with either the aromatase inhibitor anastrazole or the
ER␣ antagonist fulvestrant. The degree of lipid deposition in
both anastrazole- and fulvestrant-treated Tfm mice, however,
was still significantly lower than that of placebo-treated Tfm
mice. These data demonstrate that a component of the
response is mediated by the conversion of testosterone to
17␤-estradiol by aromatase and subsequent activation of
ER-␣. However, most of the response is mediated by testosterone acting directly through a mechanism that is indepen-
dent of the classic AR. Nathan et al7 reported that testosterone
treatment in LDL receptor knockout mice reduced atherosclerotic lesion formation but that this was not observed in mice
coadministered anastrazole, thereby implicating conversion
to 17␤-estradiol as the sole mechanism of action of testosterone in reducing lesion formation.
Long-term physiological testosterone replacement therapy
also was associated with an increase in the atheroprotective
HDL fraction of cholesterol in the Tfm mouse. Human
cross-sectional studies have demonstrated a positive association between testosterone concentrations and HDLC in both
healthy men36 and men with type 2 diabetes mellitus.37
However, in contrast to the effects of testosterone on fatty
streak formation, the effects observed on HDLC in this study
would appear to be solely attributable to the conversion of
testosterone to 17␤-estradiol. Indeed, although serum levels
of HDLC were significantly raised in Tfm mice after physiological testosterone replacement therapy alone, this effect
was not observed in Tfm mice receiving physiological testosterone replacement in conjunction with either fulvestrant
or anastrazole. Consequently, the beneficial action of testosterone in elevating HDLC levels would appear to occur by
conversion to 17␤-estradiol via aromatase and subsequent
activation of ER␣-dependent pathways. Indeed, previous
studies have reported that physiological levels of 17␤estradiol may influence serum lipoprotein concentrations,38,39
although the results of these studies are conflicting.
The nongenomic mechanisms underlying the beneficial
effect of testosterone on fatty streak formation are currently
unknown. However, a number of nongenomic mechanisms of
action of testosterone have recently been described, notably
an ability to inhibit Cav1.2 L-type calcium channels in a
manner similar to nifedipine.40 Indeed, nifedipine has been
demonstrated to exert antiatherogenic effects in cholesterolfed animal models without reducing dietary hypercholesterolemia.41 Nifedipine has been demonstrated to inhibit monocyte-macrophage–induced oxidation of LDL cholesterol
and cholesterol accumulation, thereby reducing foam cell
formation.42 Numerous studies report nongenomic effects of
testosterone in macrophages similar to nifedipine (reviewed
elsewhere13). The extravasation of monocytes into the
subendothelial cell layer is governed by several factors,
including the expression of adhesion molecules and various
cytokines. Testosterone has been demonstrated to decrease
interleukin-6, interleukin-1␤, and tumor necrosis factor-␣
production in monocyte-macrophages, as well as vascular cell
adhesion molecule-1 expression and the nuclear translocation
of nuclear factor-␬B in human endothelial cells (reviewed by
Jones et al3). In addition, testosterone therapy has been
reported to improve insulin sensitivity and to reduce elevated
plasminogen activator inhibitor-1 levels in hypogonadal men
(reviewed elsewhere3). It is likely that ⱖ1 of these mechanisms underlie the observed beneficial effects of testosterone
on fatty streak formation. Clearly, further study is warranted.
In summary, this study has demonstrated a beneficial
action of testosterone on fatty streak formation. We have
demonstrated that physiological concentrations of testosterone inhibit aortic fatty streak formation in cholesterol-fed
mice, an action that is independent of the classic AR and
Nettleship et al
Testosterone Therapy and Fatty Streak Formation
mediated in part by conversion to 17␤-estradiol. In addition,
physiological testosterone treatment increased HDLC, an
effect consistent with conversion to 17␤-estradiol and subsequent activation of genomic, ER␣-dependent pathways. These
data suggest that physiological testosterone replacement therapy
may protect against atherogenesis and potentially confer cardiovascular benefits to men with hypotestosteronemia.
Acknowledgments
The authors would like to thank Chris Biggins from Clinical
Chemistry, Royal Hallamshire Hospital, Sheffield, UK, for performing the serum cholesterol measurements and AstraZeneca for their
kind donation of Anastrazole.
Source of Funding
This research was funded by a British Heart Foundation project
grant.
Disclosures
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
None.
References
1. Bagatell CJ, Bremner WJ. Androgens in men: uses and abuses. The
N Engl J Med. 1996;334:707–714.
2. British Heart Foundation Statistics Database. Available at http://
www.bhf.org.uk. Accessed February 2007.
3. Jones RD, Nettleship JE, Kapoor DE, Jones TH, Channer KS. Testosterone and atherosclerosis in aging men. Am J Cardiovasc Drugs. 2005;
5:141–154.
4. Larsen BA, Nordestgaard BG, Stender S, Kjeldsen K. Effect of testosterone on atherogenesis in cholesterol-fed rabbits with similar plasmacholesterol levels. Atherosclerosis. 1993;99:79 – 86.
5. Bruck B, Brehme U, Gugel N, Hanke S, Finking G, Lutz C, Benda N,
Schmahl FW, Haasis R, Hanke H. Gender-specific differences in the
effects of testosterone and estrogen on the development of atherosclerosis
in rabbits. Arterioscler Thromb Vasc Biol. 1997;17:2192–2199.
6. Alexandersen P, Haarbo J, Byrjalsen I, Lawaetz H, Christiansen C.
Natural androgens inhibit male atherosclerosis: a study in castrated,
cholesterol-fed rabbits. Circ Res. 1999;84:813– 819.
7. Nathan L, Shi WB, Dinh H, Mukherjee TK, Wang XP, Lusis AJ,
Chaudhuri G. Testosterone inhibits early atherogenesis by conversion to
estradiol: critical role of aromatase. Proc Natl Acad Sci U S A. 2001;98:
3589 –3593.
8. Gordon GB, Bush DE, Weisman HF. Reduction of atherosclerosis by
administration of dehydroepiandrosterone: a study in the hypercholesterolemic New Zealand White rabbit with aortic intimal injury. J Clin Invest.
1988;82:712–720.
9. Arad Y, Badimon JJ, Badimon L, Hembree WC, Ginsberg HN. Dehydroepiandrosterone feeding prevents aortic fatty streak formation and
cholesterol accumulation in cholesterol-fed rabbit. Arterioscler Thromb.
1989;9:159 –166.
10. Eich DM, Nestler JE, Johnson DE, Dworkin GH, Ko D, Wechsler AS,
Hess ML. Inhibition of accelerated coronary atherosclerosis with dehydroepiandrosterone in the heterotopic rabbit model of cardiac transplantation. Circulation. 1993;87:261–269.
11. Hanke H, Lenz C, Spindler KD, Weidemann W. Effect of testosterone on
plaque development and androgen receptor expression in the arterial
vessel wall. Circulation. 2001;103:1382–1385.
12. von Dehn G, von Dehn O, Volker W, Langer C, Weinbauer GF, Behre
HM, Nieschlag E, Assmann G, von Eckardstein A. Atherosclerosis in
apolipoprotein E-deficient mice is decreased by the suppression of endogenous sex hormones. Horm Metab Res. 2001;33:110 –114.
13. Jones RD, English KM, Pugh PJ, Morice AH, Jones TH, Channer KS.
Pulmonary vasodilatory action of testosterone: evidence of a calcium
antagonistic action. J Cardiovasc Pharmacol. 2002;39:814 – 823.
14. Hall J, Jones RD, Jones TH, Channer KS, Peers C. Selective inhibition of
L-type Ca2⫹ channels in A7r5 cells by physiological levels of testosterone. Endocrinology. 2006;147:2675–2680.
15. Gorczynska E, Handelsman DJ. Androgens rapidly increase the cytosolic
calcium concentration in Sertoli cells. Endocrinology. 1995;136:
2052–2059.
2433
16. Guo Z, Benten WPM, Krucken J, Wunderlich F. Non-genomic testosterone calcium signaling: genotropic actions in androgen receptor-free
macrophages. J Biol Chem. 2002;277:29600 –29607.
17. Lieberherr M, Grosse B. Androgens increase intracellular calcium concentration and inositol 1,4,5-trisphosphate and diacylglycerol formation
via a pertussis toxin-sensitive G-protein. J Biol Chem. 1994;269:
7217–7223.
18. Steinsapir J, Socci R, Reinach P. Effects of androgen on intracellular
calcium of LNCaP cells. Biochem Biophys Res Commun. 1991;179:
90 –96.
19. Charest NJ, Zhou ZX, Lubahn DB, Olsen KL, Wilson EM, French FS. A
frameshift mutation destabilizes androgen receptor messenger RNA in the
Tfm mouse. Mol Endocrinol. 1991;5:573–581.
20. He WW, Kumar MV, Tindall DJ. A frame-shift mutation in the androgen
receptor gene causes complete androgen insensitivity in the testicularfeminized mouse. Nucleic Acids Res. 1991;19:2373–2378.
21. Jones RD, Pugh PJ, Hall J, Channer KS, Jones TH. Altered circulating
hormone levels, endothelial function and vascular reactivity in the testicular feminised mouse. Eur J Endocrinol. 2003;148:111–120.
22. Murphy L, O’Shaughnessy PJ. Testicular steroidogenesis in the testicular
feminized (Tfm) mouse: loss of 17-alpha-hydroxylase activity. J Endocrinol. 1991;131:443– 449.
23. Nettleship JE, Jones RD, Channer KS, Jones TH. Cholesterol feeding
induces fatty streak formation in the testicular feminised mouse. Endocr
Abstracts. 2005;9:P19.
24. English KM, Mandour O, Steeds RP, Diver MJ, Jones TH, Channer KS.
Men with coronary artery disease have lower levels of androgens than
men with normal coronary angiograms. Eur Heart J. 2000;21:890 – 894.
25. Kabakci G, Yildirir A, Can I, Unsal I, Erbas B. Relationship between
endogenous sex hormone levels, lipoproteins and coronary atherosclerosis in men undergoing coronary angiography. Cardiology. 1999;92:
221–225.
26. Phillips G, Pinkernell B, Jing T. The association of hypotestosteronemia
with coronary artery disease in men. Arterioscler Thromb. 1994;14:
701–706.
27. Pugh PJ, Channer KS, Parry H, Downes T, Jones TH. Bio-available
testosterone levels fall acutely following myocardial infarction in men:
association with fibrinolytic factors. Endocr Res. 2002;28:161–173.
28. Zhao SP, Li XP. The association of low plasma testosterone level with
coronary artery disease in Chinese men. Int J Cardiol. 1998;63:161–164.
29. Barrett-Connor E, Khaw K. Endogenous sex hormones and cardiovascular disease in men: a prospective population-based study. Circulation.
1988;78:539 –545.
30. Barrett-Connor E. Lower endogenous androgen levels and dyslipidemia
in men with non-insulin-dependent diabetes mellitus. Ann Intern Med.
1992;117:807– 811.
31. Haffner S, Mykkanen L, Valdez R, Katz M. Relationship of sex hormones
to lipids and lipoproteins in non-diabetic men. J Clin Endocrinol Metab.
1993;77:1610 –1615.
32. Barud W, Palusinski R, Beltowski J, Wojcicka G. Inverse relationship
between total testosterone and anti-oxidized low density lipoprotein
antibody levels in ageing males. Atherosclerosis. 2002;164:283–288.
33. Simon D, Charles M, Nahoul K, Orssaud G, Kremski J, Hully V, Joubert
E, Papoz L, Eschwege E. Association between plasma total testosterone
and cardiovascular risk factors in healthy adult men: the Telecom Study.
J Clin Endocrinol Metab. 1997;82:682– 685.
34. Malkin CJ, Pugh PJ, Jones RD, Kapoor D, Channer KS, Jones TH. The
effect of testosterone replacement on endogenous inflammatory cytokines
and lipid profiles in hypogonadal men. J Clin Endocrinol Metab. 2004;
89:3313–3318.
35. Kapoor D, Goodwin E, Channer KS, Jones TH. Testosterone replacement
therapy improves insulin resistance, glycaemic control, visceral adiposity
and hypercholesterolaemia in hypogonadal men with type 2 diabetes. Eur
J Endocrinol. 2006;6:899 –906.
36. Van Pottelbergh I, Braeckman L, De Bacquer D, De Backer G, Kaufman
J. Differential contribution of testosterone and estradiol in the determination of cholesterol and lipoprotein profile in healthy middle-aged men.
Atherosclerosis. 2003;166:95–102.
37. Stanworth R, Kapoor D, Channer KS, Jones TH. HDL cholesterol levels
are positively associated with testosterone and are lower with shorter
androgen receptor CAG repeat lengths in men with type 2 diabetes.
Endocr Abstracts. 2007;13:P258.
38. Shono N, Kumagai S, Higaki Y, Nishizumi M, Sasaki H. The relationship
of testosterone, estradiol, dehydroepiandrosterone-sulphate and sex
2434
Circulation
November 20, 2007
hormone binding globulin to lipid and glucose metabolism in healthy
men. J Atheroscler Thromb. 1996;3:45–51.
39. Dai W, Gutai J, Kuller L, Laporte R, Falvo-Gerald L, Caggiula A. A
relationship between plasma high-density lipoprotein cholesterol and sex
hormone concentrations in men. Am J Cardiol. 1994;53:1259 –1263.
40. Scragg J, Dallas M, Peers C. Molecular requirements for L-type Ca2⫹
channel blockade by testosterone. Cell Calcium. 2006;42:11–15.
41. Henry P, Bentley K. Suppression of atherogenesis in cholesterol-fed
rabbit treated with nifedipine. J Clin Invest. 1981;5:1366 –1369.
42. Lesnik P, Dachet C, Petit L, Moreau M, Griglio S, Brudi P, Chapman MJ.
Impact of a combination of a calcium antagonist and a ␤-blocker on celland copper-mediated oxidation of LDL and on the accumulation and
efflux of cholesterol in human macrophages and murine J774 cells.
Arterioscler Thromb Vasc Biol. 1997;17:979 –988.
CLINICAL PERSPECTIVE
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Research indicates that low testosterone rather than testosterone per se is associated with coronary artery disease in men.
Evidence also suggests that men with hypotestosteronemia with concomitant coronary artery disease may benefit from
replacement therapy. The mechanisms by which testosterone produces these beneficial effects have yet to be elucidated.
The testicular feminized (Tfm) mouse exhibit a nonfunctional androgen receptor (AR) and low circulating testosterone
concentrations. We used the Tfm mouse to determine whether testosterone modulates atheroma formation via its classic
signaling pathway involving the nuclear AR, conversion to 17␤-estradiol, or an alternative signaling pathway. This study
demonstrates that low endogenous testosterone, be it associated with a nonfunctional AR (in the Tfm mouse) or a fully
functional AR (in the castrated XY male), is associated with fatty streak formation within the aortic root after feeding on
cholesterol-enriched diet. Physiological testosterone replacement therapy was shown to prevent aortic fatty streak
formation in the Tfm mouse and to raise high-density lipoprotein cholesterol. After testosterone replacement therapy in
conjunction with either anastrazole or Faslodex, however, this protective effect was reduced although still significantly
lower than that of placebo-treated Tfm mice. Improvement in high-density lipoprotein cholesterol also was completely
abolished by cotreatment with these agents. These data provide evidence of a dual modulating effect of testosterone and
17␤-estradiol in providing a beneficial action on fatty streak formation. Physiological testosterone replacement inhibited
aortic fatty streak formation in cholesterol-fed mice, an action that was independent of the classic AR. Additionally,
physiological testosterone replacement increased high-density lipoprotein cholesterol, an effect consistent with conversion
to 17␤-estradiol and subsequent activation of genomic, estrogen receptor ␣– dependent pathways.
Physiological Testosterone Replacement Therapy Attenuates Fatty Streak Formation and
Improves High-Density Lipoprotein Cholesterol in the Tfm Mouse: An Effect That Is
Independent of the Classic Androgen Receptor
Joanne E. Nettleship, T. Hugh Jones, Kevin S. Channer and Richard D. Jones
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Circulation. 2007;116:2427-2434; originally published online November 5, 2007;
doi: 10.1161/CIRCULATIONAHA.107.708768
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2007 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/116/21/2427
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation 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 is online at:
http://circ.ahajournals.org//subscriptions/