?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
Soy diet worsens heart disease in mice
Brian L. Stauffer,1,2 John P. Konhilas,2 Elizabeth D. Luczak,2 and Leslie A. Leinwand1,2
1Department of Medicine, Division of Cardiology, University of Colorado Health Sciences Center, Denver, Colorado, USA. 2Department of Molecular,
Cellular, and Developmental Biology, University of Colorado at Boulder, Boulder, Colorado, USA.
The promise of nutritional genomics in considering the pathogen-
esis and treatment of disease is beginning to be recognized (1–5).
For example, soy-rich diets are perceived to be generally beneficial
from a health standpoint, particularly in the context of the cardio-
vascular system (6). This is exemplified by the fact that American
consumers will spend $4.7 billion on soy foods and dietary sup-
plements. However, there has not been a systematic experimental
evaluation of how individuals with defined genetics respond to dif-
ferent diets and how sex might modify this interaction. One start-
ing point to evaluate the effect of a particular diet such as soy is
the laboratory rodent, where the standard diet is soy-based. Potent
physiologic effects of soy diets have been demonstrated in both
laboratory and clinical settings. Some of these effects have been
reported to be beneficial and some may have adverse consequences.
Among the beneficial properties are the prevention of cancer and
a lowering of cholesterol (7). Among the potentially adverse effects
are an increase in androgen levels and a decrease in thyroid peroxi-
dase (8, 9). Many of the physiologic effects of a soy diet have been
attributed to the soy isoflavones or phytoestrogens, and there are
many experimental studies that have studied the effects of genistein
and daidzein, the 2 most prominent isoflavones in soy (10). Indeed,
soy isoflavones improve hyperlipidemia and cardiovascular disease
associated with abnormalities in lipid metabolism via activation of
PPARα (11, 12). In human and rodent heart failure, PPARα activity
is attenuated (13, 14). However, a recent large epidemiologic study
revealed no beneficial effect of dietary phytoestrogens on the inci-
dence of clinical coronary or cerebrovascular events in women (15).
Despite this focus on phytoestrogens, soy has many complex nutri-
ents and is consumed as a major part of the diet in many cultures.
Further, decades of literature on experimental laboratory rodents
have been in the context of a soy-based diet (rather than a diet sup-
plemented by dietary phytoestrogens), and thus it is of great inter-
est to examine the interaction of the soy diet and genetic models of
disease. In the current study, we have asked what the impact of diet
is on a genetic mouse model of cardiac hypertrophy and how sex
might modify such a proposed interaction.
Myocardial hypertrophy in response to a disease stimulus con-
sists initially of compensatory myocellular enlargement. However,
the heart ultimately reaches a point where the stress overwhelms
compensatory processes, and ventricular chamber enlargement,
wall thinning, and impaired contractile function ensue, leading to
heart failure. The myocardial mechanisms underlying this transi-
tion are thus far unknown.
Sex-specific variation in myocardial hypertrophy and progres-
sion to heart failure have been clearly documented over the past
several decades (16, 17). In response to a disease stimulus (hyper-
tension, ref. 18; valvular disease, ref. 17; sarcomeric mutations, ref.
19; and aging, refs. 20, 21), both sexes initially develop LV hyper-
trophy. However, men subsequently develop heart failure with
chamber dilation, wall thinning, and impaired contractile func-
tion. Women develop heart failure with preserved LV contractile
function. The sex difference in cardiac function favors the survival
of women with heart failure (22, 23).
While these sex-differences in clinical cardiac disease are reason-
ably well characterized, there is a dearth of mechanistic informa-
tion about the triggers of increased mortality. Several studies have
evaluated the effects of sex hormones (24, 25) and phytoestro-
gens (15) on the cardiovascular system in humans, predominant-
ly investigating effects on heart attack or stroke, disorders more
associated with abnormalities of blood vessels. While estrogen
and testosterone receptors have been identified in the myocar-
dium (21, 26, 27), less is known about the influence of sex hor-
mones on the myocardium.
We have developed a genetic mouse model of hypertrophic car-
diac disease that exhibits the sex-dependent phenotypic charac-
teristics documented in clinical populations (28). That is, while
females preserve cardiac contractile function and continue to
increase their cardiac mass, male mice develop thin ventricular
walls and have poorly contractile hearts. This hypertrophic cardio-
myopathic (HCM) mouse expresses a mutant myosin heavy chain
(MyHC) transgene in the heart, and all of the characterization has
been performed on animals fed a soy diet (29). We hypothesized
that phytoestrogens influence cardiac growth and that the stan-
dard rodent diet plays an influential role in the development of
this dilated, poorly contractile phenotype in male mice. We tested
this hypothesis by comparing HCM mice and WT littermate con-
trols of both sexes consuming a standard soy-based diet, a casein-
based diet (30, 31), or a casein diet supplemented with daidzein
and genistein. Further, we determined the effect of diet on several
aspects of known hypertrophic signaling.
Nonstandard?abbreviations?used: BW, body weight; ER, estrogen receptor; GSK3β,
glycogen synthase kinase 3β; HCM, hypertrophic cardiomyopathy; HW, heart weight;
MyHC, myosin heavy chain; TL, tibia length; VW, ventricular weight.
Conflict?of?interest: The authors have declared that no conflict of interest exists.
Citation?for?this?article: J. Clin. Invest. 116:209–216 (2006).
Related Commentary, page 16
210?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
Diet and cardiac growth. Diet was found to significantly affect car-
diac mass. Ventricular weight (VW) (RV and LV) was normalized to
tibia length (TL) (Figure 1A). As expected, the HCM animals had
higher VW/TL than their sex-matched WT controls fed the same
diet (P < 0.005 for all 4 groups). Also, as expected, male animals
had larger hearts than female animals (P < 0.0001 for all 4 groups).
However, HCM animals had significantly more hypertrophy on
the casein diet than on the soy diet (P < 0.05 each). The male HCM
animals fed a casein-based diet had greater cardiac hypertrophy
(17%) than the HCM males fed a soy diet (9%) when compared with
control mice. In contrast, diet did not have an effect on hypertro-
phy in female HCM animals (Figure 1B). As might be expected,
diet also influenced body weight (BW). All mice consuming the
casein diet were significantly heavier than those on the soy diet.
Body composition, evaluated by dual-energy x-ray absorptiometry
(DEXA), demonstrated no significant diet-dependent differences
in lean BW; however, the animals consuming the casein diet had
higher body fat than those consuming the soy-based diet. Impor-
tantly, normalization of VW for BW did not influence the differ-
ences between the transgenic groups noted above.
Dietary estrogens and cardiac growth. There are several components
that are different between the soy and casein diets. We investigated
the possibility that phytoestrogens contribute to differences in
cardiac growth between the diets. The phytoestrogens daidzein
and genistein were added to the casein diet. This supplemented
diet was fed to male and female WT and HCM mice and compared
with mice consuming the standard soy and casein diets. There was
no significant effect of diet on heart weight/BW (HW/BW) ratios
in either the male or female WT mice at 2 months of age. However,
there was a significant effect of diet on HW/BW ratio in the male
and female HCM animals at this early time point (P < 0.01). More-
over, the influence of diet was observed in a sex-dependent fashion
(sex × diet interaction, P < 0.05). The influence of adding phytoes-
trogen to the casein diet was greater in the male mice, augment-
ing the HW/BW ratio by approximately 6%, while this addition
attenuated cardiac growth in the female HCM mice by 6% (Figure
2). Importantly, cardiac growth was lower in the HCM animals of
both sexes consuming the standard soy diet, and the addition of
the phytoestrogens to the casein diet did not completely recapitu-
late this phenotype.
Endogenous sex steroids and cardiac growth. The influence of the
dietary compounds may occur through estrogenic effects on
endogenous sex steroid receptors. These receptors have been
implicated in cardiac growth in other models (32, 33). We inves-
tigated the possibility that diet contributes to the phenotype via
a sex steroid receptor–mediated mechanism. Male and female
HCM mice on the casein diet underwent prepubertal surgical
gonadectomy with subsequent placebo, estrogen, or testosterone
supplementation or sham operation. Gonadectomy with placebo
supplementation (absence of all sex steroids) resulted in lower
cardiac mass regardless of sex compared with sham-operated
animals (males: 3.37 ± 0.19 mg/g BW versus 4.16 ± 0.37 mg/g;
females: 2.93 ± 0.21 mg/g versus 3.45 ± 0.90 mg/g). Testosterone
supplementation of the females (4.47 ± 0.10 mg/g) and estrogen
supplementation of the males (3.93 ± 0.27 mg/g) caused cardiac
mass to be similar to the opposite sex sham-operated animals. For
comparative numbers, see those noted above. In summary, lack of
sex hormones results in lower HW/BW ratios in both sexes, and
testosterone and estrogen appear to be primary determinants of
normalized heart mass.
Diet and in vivo cardiac function. Cardiac contractile function evalu-
ated by echocardiography was profoundly affected by diet in male
HCM animals (Figure 3). At 8 months of age, HCM males fed a soy
diet had significantly depressed contractile function (P < 0.0001)
while animals fed a casein diet were indistinguishable from WT.
Cardiac growth due to transgene expression is modified by sex and
diet at 8 months of age. (A) VW normalized to body growth, assessed
by TL. *Significant transgene effect with each sex and diet combination
(P < 0.005 for each). †Significant sex effect by diet in both WT and HCM
mice (P < 0.0001). #Significant diet effect (P < 0.05) between groups
except in WT males. (B) There is a diet-dependent difference in the
degree of hypertrophy in the male mice while there is no diet-depen-
dent difference in the female animals. §P < 0.01 versus male casein.
n = 21–35 in each group. HCM, transgenic HCM mice. Mean ± SEM.
Cardiac growth due to diet in male and female HCM mice consum-
ing a standard soy diet, a casein, phytoestrogen-free diet, or a casein
diet supplemented with the phytoestrogens daidzein and genistein (PE
diet) was found to occur in a sex-dependent fashion. ANOVA main
effects: diet, P < 0.0001; sex × diet, P < 0.05. n = 12–19 in each group
except in those on the phytoestrogen-supplemented diet (n = 6).
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
The decrease in contractile function (Figure 3A) was accompanied
by significant wall thinning (Figure 3B) and LV chamber dilation.
Thus, a soy-based diet has a very significant detrimental effect on
cardiac function in this model of hypertrophic disease.
Diet and blood pressure in vivo. Estrogen is associated with improve-
ments in endothelium-mediated arterial vasodilation, a marker of
vascular health. Oral genistein supplementation has also been
shown to improve endothelial function in humans (5). Altera-
tions in endothelial and arterial function may lead to alterations
in blood pressure that influence the development of cardiac hyper-
trophy. We therefore sought to determine whether the sex- and
diet-dependent differences were due to differences in systemic
blood pressure. In vivo hemodynamics with Millar catheterization
revealed no significant differences in mean, systolic, or diastolic
arterial pressure (data not shown). The lack of influence of sex or
diet on blood pressure suggests a primary myocardial influence of
diet on the differences observed in cardiac function.
Diet and myocardial fibrosis and myocellular disarray. The effect of
diet on histopathology as evidenced by collagen deposition and
myofilament disorganization was evaluated at 8 months of age
using picrosirius red staining. Fibrosis and myocellular disarray
can be visualized under polarization light microscopy by yellow/
orange birefringence and lack of green sarcomeric birefringence,
respectively. Highly ordered sarcomeres have light-green birefrin-
gent properties (Figure 3). Increased collagen deposition causes
increased myocardial stiffness and is associated with myocellular
disarray. Both of these features have been implicated in myocardial
contractile dysfunction (34–36). HCM animals display an increase
in collagen deposition and myocellular disarray compared with
WT (Figure 4). This was evidenced by a transition from predomi-
nantly thin collagen fibrils (green birefringence) in the WT con-
trols to predominantly thick collagen fibers (yellow/orange bire-
fringence) in the HCM male and female mice (Figure 4 and data
not shown). The lack of uniform myofilament birefringence in the
HCM sections is also indicative of an increase in myocellular disar-
ray. Most importantly, there was a distinctly worse histopathologic
phenotype in the soy HCM males when compared with the casein
HCM males. We observed no significant difference in fibrosis and
disarray among diets in the female HCM mice. Moreover, the his-
topathologic phenotype was indistinguishable between the sexes
on the casein diet. The increased fibrosis and myocellular disarray
observed in the male HCM mice consuming the soy-based labora-
tory diet seem likely to contribute to the contractile dysfunction
in this experimental group.
Diet and β-MyHC protein synthesis. Heart failure in rodents and
humans is associated with increased levels of the slower, more
energy efficient myosin motor protein, β-MyHC. α-MyHC, the
predominant isoform in murine hearts, declines with worsening
heart failure and is accompanied by increases in β-MyHC (37). An
increase in β-MyHC protein content acts in a dominant fashion
and is one of the mechanisms underlying myocardial contractile
dysfunction (38, 39). Additionally, β-MyHC content can be used
as an objective assessment of disease severity. We measured the
β-MyHC content of the LV at 8 months of age (Figure 5, A and
B). Because contractile function was abnormal in the male HCM
mice consuming the soy diet, we hypothesized that more β-MyHC
protein would be observed in this group. Indeed, male HCM mice
fed the soy diet expressed more β-MyHC than the female HCM
mice, consistent with the more severe disease state in males. In
addition, there was strikingly lower β-MyHC content in the male
HCM mice consuming the casein diet. This is of particular interest
since the absolute and normalized heart weights were greater in
the casein-fed mice. This suggests that the growth that occurred
in the HCM hearts due to the casein diet is a beneficial (or physi-
ologic) hypertrophy. The increased β-MyHC protein expression
may be a mechanism contributing to the impairment in systolic
function in the male HCM mice on the soy diet.
Diet and the IGF-1/Akt/glycogen synthase kinase 3 protein kinase cas-
cade. Since IGF-1 treatment of cardiac myocytes in culture increases
β-MyHC synthesis (40, 41), we explored the IGF-1/Akt/glycogen syn-
thase kinase 3 (IGF-1/Akt/GSK3) cascade. Sex-dependent differences
In vivo cardiac function assessed by echocardiography in HCM mice
at 8 months of age. (A) Decreased contractile function (% fractional
shortening) is observed in male HCM mice consuming the soy diet.
The impairment is ameliorated on the casein diet. (B) The functional
impairment is associated with thinning of the anterior wall of the LV.
These abnormalities are prevented by changing the animals to a phy-
toestrogen-free diet. n = 5–13 in each group. *P < 0.0001 versus male
casein; †P < 0.05 versus female casein; #P < 0.05 versus same diet,
alternate sex. Mean ± SEM.
Cardiac fibrosis is a mechanism behind contractile dysfunction in heart
failure. An improvement in cardiac fibrosis is observed in the male HCM
mice consuming a casein diet. n = 5 in each group. Increased collagen
deposition in the male HCM mice on the soy diet may contribute to the
contractile dysfunction observed by echocardiogram.
212? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
in the IGF-1/Akt/GSK3β pathway have been observed in humans
and in animal models of cardiac disease (42). Several animal models
have indicated that this pathway is involved in cardiac hypertrophy
via phosphorylation of Akt (protein kinase B; activation) and GSK3β
(inactivation) (43, 44). Increased nuclear localization of phospho-Akt
has been observed in the myocardium of premenopausal women
as well as in response to estrogen stimulation in cultured cardio-
myocytes (42). IGF-1 has been identified as one of several potential
ligands responsible for estrogen-independent activation of estrogen
receptors (ERs) (45, 46). Based on this information, we evaluated
GSK3β, a terminal protein in this cascade, for changes in activity
as a potential protein kinase leading to the sex and diet differences
in this disease model. GSK3β in the active, unphosphorylated form
suppresses protein synthesis by inhibiting translation initiation fac-
tor eIF2b. Phosphorylation of GSK3β allows translation initiation
and promotes hypertrophy in cardiac myocytes (47).
We hypothesized that the soy diet increases activity of the IGF-1/
Akt/GSK3β cascade, which is then permissive of β-MyHC pro-
tein synthesis. Therefore, the phosphorylation state of GSK3β as
a fraction of total GSK3β was measured in LV homogenates. As
predicted, the HCM mice exhibited a 2- to 3-fold higher ratio of
phosphorylated (inactive) to total GSK3β compared with WT con-
trols (Figure 5C). However, phosphorylated GSK3β was over 3-fold
higher in the male HCM mice fed a soy diet compared with male
HCM mice fed a casein diet. This elevation correlates well with
the significant increase in β-MyHC protein. While we expected to
observe a higher ratio of phospho- to total GSK3β in the HCM mice,
the diet difference in the HCM males is an unexpected and novel
finding. Significantly higher phosphorylation states as observed in
the HCM males consuming the soy diet may be a mechanism con-
tributing to the transition to the decompensated state.
Diet and apoptosis. One mechanism underlying cardiac dilation
is myocellular apoptosis. Sex differences in apoptotic myocardial
cell loss have been identified (20). Importantly, genistein induces
apoptosis via caspase-3 in a number of settings (48). For example,
in human prostate cancer cell lines, genistein induces apoptosis
through expression and activation of caspase-3 (49). Caspase-3
has been shown to be a proapoptotic factor in cardiac myocytes
by initiating mitochondrial protein dissociation (50). An increase
in programmed myocellular death therefore could underlie the
development of the dilated phenotype in the male HCM mice on a
soy diet. We measured caspase-3 activity in the hearts at 8 months
of age. There is a marked elevation in caspase-3 activation in the
male HCM mice consuming the soy-based diet, which is reversed
in male HCM mice consuming a casein-based diet (Figure 6).
The salient findings from the current study are as follows: (a)
Diet significantly alters cardiac structure and function in WT and
HCM animals; (b) The diet-dependent phenotype differs signifi-
cantly between sexes at the macroscopic and cellular levels; (c) The
phytoestrogens daidzein and genistein influence cardiac growth
in a sex-dependent fashion; (d) Diet-dependent alterations in the
Positive adaptation is observed in male mice consuming a casein diet.
β-MyHC exerts a dominant effect in heart failure and is a mechanism
behind impairments in contractile function. (A) β-MyHC (expressed
as a percentage of total MyHC content) is increased in hearts from
8-month-old mice consuming a soy diet. n = 3–7 in each group. *P < 0.05
versus WT same sex and diet and same sex, alternate diet HCM mice.
†P < 0.05 versus alternate sex, same diet HCM mice. (B) Representa-
tive silver-stained MyHC separation gel from WT male and HCM male
mice. (C) Increased activation of the IGF-1 receptor upregulates MyHC
protein in the heart. GSK3β is a protein kinase downstream from the
IGF receptor. The ratio of phospho- to total GSK3β is significantly
attenuated in the males consuming a phytoestrogen-free diet. The
increased activity down this cascade may be permissive of increased
β-MyHC protein synthesis. n = 5 in each group. **P < 0.05 versus WT
same sex and diet; #P < 0.05 versus alternate diet, same sex HCM
mice and alternate sex HCM mice regardless of diet. Mean ± SEM.
Increased caspase-3 activity is indicative of increased myocellular
apoptosis and may be a mechanism contributing to cardiac dilata-
tion in male HCM mice consuming the soy diet. Caspase-3 activity
is significantly attenuated in HCM males consuming the casein diet.
n = 5–13 in each group. *P < 0.05 versus WT same sex and diet;
†P < 0.05 versus alternate diet, same sex HCM mice and alternate sex
mice regardless of diet. Mean ± SEM.
? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
IGF-1/Akt/GSK3β pathway, β-MyHC content, and caspase-3 acti-
vation are potential molecular mechanisms responsible for these
changes in phenotype. To our knowledge, this is the first report of
significant differences in cardiac muscle adaptation due to dietary
manipulation. These data strongly suggest further investigation
into a link between diet and cardiomyopathy.
A major difference between the 2 diets in the current study is the
protein source. The standard laboratory diet is soy based, contain-
ing significant amounts of phytoestrogens, which are major steroid
receptor ligands and have been shown to have multiple biological
effects. Although the relative estrogenic activity is 10–2- to 10–3-fold
less than estradiol or estrone, the principal circulating estrogens in
most mammals (4), there are several characteristics that promote
biological activity. These estrogenic compounds have less nonspe-
cific binding to proteins in blood, which enhances the number of
molecules available for receptor binding (51). In addition, serum
concentrations of phytoestrogens with modest dietary intake have
been observed to be 104- to 105-fold higher than endogenous estra-
diol concentrations (4). Phytoestrogens also have a preference for
the ERβ and several non-ER related actions (i.e., protein tyrosine
kinase inhibitor and antioxidant). Perhaps most importantly, these
pleiomorphic effects have been observed to occur within an indi-
vidual organ or cell, consistent with steroid action in general.
We and others (52–56) have demonstrated that sex steroids
influence cardiac growth. The exact mechanism of action is
unclear, but additional investigation is ongoing. These results
establish the biological plausibility of the influence of phytoestro-
gens on the heart to elicit cardiac growth via a sex steroid–receptor
mechanism. While there are several reasonable theories regarding
the influence of phytoestrogens on sex hormone receptors in the
heart, we would propose that it is a difference in the effective estro-
gen dose between the sexes that may be responsible for the results
of our investigation.
Since sex-dependent alterations in myocardial ERs have never
been described, we would propose that the augmented estrogenic
exposure in the male mice contributes to the dilated phenotype.
We speculate that, because female animals have higher endog-
enous estrogen levels, the proportional increase in estrogenic
compounds via diet is less in females compared with males, who
are chronically exposed to significantly lower levels of estrogenic
compounds. For example, it has been estimated that basal female
17β-estradiol levels are 5–30 pg/ml while males have negligible
levels (57–60). Serum phytoestrogen levels have been reported in
the 1000–2000 ng/ml range on several standard rodent diets inde-
pendent of sex (61).
Differences in affinity and estrogenic activity between phytoes-
trogens and endogenous estrogens make it difficult to assess the
proportion of biological activity due to each individual compound
in female animals. In males this is not the case since endogenous
estrogenic compounds are present only in negligible quantities;
therefore the full effect of the phytoestrogens would be observed.
To support this argument further, Boettger-Tong et al. have dem-
onstrated a failure of the usual biochemical and morphological
uterine responses to exogenously administered estradiol in ani-
mals that were fed a rodent diet with soy meal and alfalfa (62). This
occurred when the rodent diet had been reformulated without any
investigator notification. Since, as noted above, isoflavone levels
have not been demonstrated to be different between males and
females on identical diets, the absolute influence of these com-
pounds should be substantially greater in the male animals that
have virtually no basal estrogenic stimulation. Moreover, there
is in vivo evidence that genistein acts as an agonist at ERs when
administered alone and as an antagonist when coadministered
with 17β-estradiol (63). It is plausible that a similar mechanism is
behind the sex difference in cardiac phenotype.
Indeed, our results demonstrate that phytoestrogen supplemen-
tation augments cardiac growth in the HCM males and attenuates
growth in the HCM females. The current results do not make any
assumptions regarding the molecular and functional phenotype
of the HCM mice on the supplemented diet. Importantly, adding
phytoestrogens to the casein diet does not recapitulate the entire
cardiac phenotype of the young animals consuming the soy diet.
This is not surprising given that the dietary milieu is complex and
unregulated in the soy diet. There are clearly other components in
the diet that influence cardiac growth. This fact accentuates the
importance of dietary intake in other experimental models. How-
ever, it is beyond the scope of the current study to differentiate the
other components of the soy diet that may contribute to the sex
difference in phenotype.
Phytoestrogens also activate PPARα, which is downregulated in
heart failure in vivo. In our genetic model, we observed increased
cardiac growth in animals consuming a soy-free diet, suggesting
that PPARα is not contributing to the improvement in the phe-
notype in our model. However, the casein diet was also associated
with an improvement in cardiac function assessed by echocar-
diography in the absence of a change in systemic blood pressure.
Importantly, the increased growth was accompanied by improved
cellular markers of disease including lower β-MyHC expression
and improved cellular architecture (less thick collagen deposition
and less myocellular disarray).
The soy-based laboratory diet was associated with increased
caspase-3 activity and increased activity through the IGF-1/Akt/
GSK3β pathway in the diseased animals relative to WT. IGF-1 stim-
ulation is associated with increased β-MyHC synthesis in rat models
(64). Augmentation of this pathway may lead directly to the systolic
impairment observed in the male HCM mice on the soy diet. Alter-
natively, the increased activity along this pathway may be a second-
ary phenomenon in response to the differences in cardiac growth.
Future directions will include defining the role of this pathway by
crossbreeding the HCM model with a constitutively active GSK
transgenic mouse (44) and identification of a common upstream
activator. The increase in caspase-3 activity in HCM hearts on a soy
diet indicates elevated levels of myocellular death, or apoptosis. Pre-
liminary data from our laboratory (data not shown) demonstrate
that the increase in caspase-3 activity was associated with decreased
levels of Bcl-2 (an inhibitor of apoptosis) and pro–caspase-9 (a pre-
cursor of effector caspases). These additional preliminary results
further support the conclusion that there is augmented apoptosis
in the HCM mice on a soy diet. Genistein has been shown to induce
apoptosis via activated caspase-3 in a number of settings (48, 49).
An increase in apoptosis is likely an important contributor to the
adverse remodeling observed in dilated cardiomyopathy in human
clinical populations and may be responsible in part for the tran-
sition from the compensated hypertrophic to the decompensated
dilated state in the mouse model.
Our results suggest that the dilated phenotype observed in the
male HCM mice on the soy diet is a result of an augmented growth
pathway and an augmented programmed cell death pathway. It
is likely that the decompensated phenotype results from a transi-
tion from a balance to an imbalance of cell growth and cell death.
214? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
Components of the soy diet have been associated with both myo-
cellular hypertrophy and apoptosis in several other model systems
and may be modifiers of the balance between these 2 processes
in vivo. Investigation of functional and molecular changes occur-
ring in the hearts of mice on the phytoestrogen-supplemented diet
will further elucidate the influence of phytoestrogens on cardiac
disease. Additional investigation into other specific components
of the soy diet as well as identification of characteristics of these
pathways during the compensated state are necessary to fully elu-
cidate these mechanistically.
Taken together these data show an increase in nonpathologic,
or physiologic, cardiac growth in this HCM mouse model on a
soy-free diet. The modifiers of caspase activity and the IGF-1/Akt/
GSK3β hypertrophic pathway are unknown but deserve additional
attention. This discovery may lead to novel dietary modifications
that prevent progression or formation of some hypertrophic cardi-
ac diseases. Future studies will be directed against identifying spe-
cific genetic modifiers that affect the response of the heart to diet.
These will include testing the effect of diet on different strains of
inbred mice. There is a large literature of human studies in which
correlations have been made between certain diets and diseases,
particularly cancer (3). And, in some cases, the results have been
inconclusive. The kind of study reported here on genetically identi-
cal groups of mice of both sexes should address the issue of genetic
variability in the human studies and begin to define the role that
genes play in dietary responses.
Animals. The HCM mouse model used in this study expressed a mutant rat
α-MyHC with expression driven by an α-MyHC promoter (29). The trans-
gene coding region contained 2 mutations, a point mutation, R403Q, and a
deletion of 59 amino acids in the actin binding site bridged by the addition
of 9 nonmyosin amino acids. Male and female WT and HCM mice were fed
a soy diet, a casein (phytoestrogen-free) diet, or a casein diet supplement-
ed with phytoestrogens ad libitum. Mice were genotyped by PCR from a
2-mm tail sample prior to study inclusion. Given the large number of animals
included in this study, repeat genotyping was performed on a separate tissue
sample at the time of animal sacrifice to assure correct grouping. There were
no misidentified animals noted. All animal protocols were approved by the
Institutional Animal Care and Use Committee at the University of Colorado
at Boulder and at the University of Colorado Health Sciences Center.
Diets. The formulation of the casein diet (D10001; Research Diets) fol-
lows the guidelines recommended by the American Institute of Nutrition
in 1977 (65). The soy diet used in the present study was the Harlan Teklad
Sterilizable Rodent Diet (8656; Harlan Teklad). The casein and soy diets
are composed of approximately the same percentage of protein (20–24%),
fat (4–5%), carbohydrate (65%), and fiber (5%) with similar amounts of vita-
mins and minerals. Differences between the 2 diets, however, include the
source of protein (casein in the D10001 versus alfalfa and soybean in the
8656 diet), carbohydrate (sucrose and corn starch in the D10001 versus
corn and soybean meal in the 8656 diet), and fat (corn oil in the D10001
versus soybean oil in the 8656 diet). The effective caloric content was differ-
ent between the diets as well (D10001, 3.9 kcal/g versus 8656, 3.25 kcal/g).
However, total caloric consumption was similar between groups since the
animals consuming the soy diet consumed more food (approximately
10% more). The isoflavone concentration of the soy-based rodent diet was
determined at 206 mg isoflavone/kg dry food weight daidzein and 229.5
mg isoflavone/kg dry food weight genistein. The phytoestrogen supple-
mentation diet was created by adding comparable amounts of daidzein
and genistein (LC Laboratories) to the casein diet.
Surgical gonadectomy. At 1 month of age, mice were anesthetized with 2.5%
tribromoethanol (Avertin) by intraperitoneal injection. Females under-
went bilateral oophorectory through bilateral paraspinal incisions. Males
underwent orchiectomy via a transverse scrotal incision. The incisions were
repaired with suture and skin staples. While still under anesthesia, sus-
tained release (90 day) hormone pellets (Innovative Research of America)
of 17β-estradiol, testosterone, or vehicle placebo were implanted subcu-
taneously in the nape of the neck. Sham-operated animals were handled
identically except the gonads were not removed.
Gravimetric measurements. At 2 or 8 months of age, mice were euthanized
using approved methods. Each mouse was weighed, and the length of the
right tibia was recorded. Hearts were rapidly excised and washed with ice-
cold normal saline (9% NaCl w/v). The great vessels and all atrial tissue
were removed under a dissecting scope (Carl Zeiss), and the ventricles were
blotted dry and weighed. The RVs and LVs were rapidly separated and indi-
vidually flash frozen in liquid nitrogen. The samples were stored at –80°C
until further analysis was performed.
Echocardiography. Echocardiography was performed on animals at 8
months of age as previously described (66). A VingMed System Five echo-
cardiography machine (GE Medical Systems) with a 10-MHz–phased
array transducer was used for digital image acquisition. The mice were
positioned prone, and M-mode recordings were obtained and saved on
magnetic optical media for offline analysis, which was performed using
EchoPAC software (version 6; GE Medical Systems). Fractional shortening
was calculated using the average over 3 cardiac cycles.
Blood pressure measurement. In vivo systemic blood pressure measurement
was performed on animals at 8 months of age. Each mouse was injected
intraperitoneally immediately before evaluation with 2.5% tribromoetha-
nol (Avertin). A longitudinal skin incision was performed over the trachea,
and the right carotid artery was isolated using blunt dissection under ×6.5
magnification (Carl Zeiss). The cephalad portion of the carotid artery was
ligated at the base of the skull. A small nick was made in the midportion of
the artery, and a 1.4-fr Millar Mikro-tip pressure transducer (Millar Instru-
ments) was advanced through the nick and into the aorta. After a period
of equilibration, aortic pressures were recorded. Following blood pressure
measurements, the animals were euthanized.
Histology. Following euthanasia at 8 months of age, hearts were rapidly
excised and washed, and whole-heart weight was recorded. The entire heart
was placed in 10% neutral-buffered formalin for 24 hours for fixation prior
to histological evaluation. The fixed hearts were processed, embedded in
paraffin, sectioned, and stained with picrosirius red according to standard
protocols. Thin collagen fibers showed green birefringence, and thick col-
lagen fibers showed bright yellow/orange birefringence under polarization
light microscopy. Myofibrils in series exhibited slight birefringence.
Western blot analysis. Frozen LV tissue was homogenized on ice in a protein
extraction buffer: NaCl (137 mM); Tris(hydroxymethyl)-aminomethane
(20 mM); 10% vol/vol glycerol; 1% vol/vol NP-40, pH 7.4. The homogenized
tissue was centrifuged at 12,000 g for 10 minutes at 4°C and the superna-
tant removed. Protein concentration was determined using the Bradford
method. SDS-PAGE separation was performed under denaturing condi-
tions. The proteins were transferred to a PVDF membrane (Amersham
Biosciences) using standard techniques. The membranes were probed with
antibodies (Santa Cruz Biotechnology) specific for either total GSK3β or
the phosphorylated isoform. Immunoreactivity was visualized using a
Western Lightning chemiluminescence detection system (PerkinElmer)
and quantified using densitometry.
MyHC expression. LV tissue was homogenized in a myosin sample buffer:
urea (8 mol/l); thiourea (2 mol/l); Tris, pH 6.8 (0.05 mol/l); dithiothrei-
tol (DTT) (0.075 mol/l); 3% SDS. Each sample was heated at 100°C for 3
minutes and then loaded in 6 wells containing escalating amounts of total
?The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
protein (0.5 to 50 µg/well). The samples were separated by 6% SDS-PAGE at
16 mA for 4.5 hours at 8°C. Gels were silver stained, and the percentages of
α- and β-MyHC protein were determined by densitometry (67).
Caspase-3 activity assay. Caspase-3 activity was determined by moni-
toring the rate of cleavage of a fluorogenic caspase-3 specific substrate
(Acetyl-AspGluValAsp-AMC; Calbiochem). To do this, frozen hearts
were mechanically disrupted in an ice-cold lysis buffer (0.02 ml/g tissue):
Tris(hydroxymethyl)-aminomethane (20 mM); NaCl (137 mM); EDTA
(0.2 mM); EGTA (0.5 mM); Triton X-100 (1%); glycerol (10%), pH 7.4.
In a 96-well plate, 0.5 mg of protein was added to each well with an
equal volume (50 µl) of caspase-3 activity assay buffer containing:
Tris(hydroxymethyl)-aminomethane (50 mM); EDTA (0.5 mM); glycerol
(20%); caspase-3 substrate (0.02 mM); DTT (0.004 mM), pH 7.0. Cleavage
of the substrate was monitored by excitation at 380 nm and emission at
460 nm with a Fluorskan Ascent Microplate Fluorometer (Thermo Elec-
tron Corp.). Caspase-3 activity was determined by calculating the slope
of the linear portion of the cleaved substrate and then normalized to
protein content (fluorescent units/min/mg protein).
Data and statistical analysis. Results are presented as mean ± SEM. Between
group differences of morphometric, protein, and blood pressure, data were
determined using a 2 between (male, female) by 2 between (soy, casein) by
2 between (WT, HCM) ANOVA at 8 months of age. Between group dif-
ferences of echocardiographic indices were determined using a 2 between
(male, female) by 2 between (soy, casein) ANOVA. Dietary between group
differences were determined using a 2 between (male, female) by 3 between
(soy, casein, supplemented) ANOVA. Interactions and/or specific main
effects differences were examined by Tukey post hoc analyses. P < 0.05 was
considered significant a priori.
This work was supported by grants from the NIH (2R01HL050560
to L.A. Leinwand; 5F32HL070509 to J.P. Konhilas; and
1F32HL67543 to B.L. Stauffer) and the American Heart Associa-
tion (0120679Z to B.L. Stauffer). Isoflavone concentrations of the
soy-based rodent diet were kindly determined by Patricia A. Mur-
phy, Iowa State University, Ames, Iowa.
Received for publication February 8, 2005, and accepted in revised
form October 18, 2005.
Address correspondence to: Leslie A. Leinwand, Molecular, Cellu-
lar, and Developmental Biology, Room A417, UCB 354, Boulder,
Colorado 80309, USA. Phone: (303) 492-7606; Fax: (303) 492-
8907; E-mail: Leslie.Leinwand@colorado.edu.
1. Ordovas, J.M., and Corella, D. 2004. Nutritional
genomics. Annu. Rev. Genomics Hum. Genet. 5:71–118.
2. Grierson, B. 2003. What your genes want you to eat.
The New York Times Magazine. May 4. 76–77.
3. Glazier, M.G., and Bowman, M.A. 2001. A review
of the evidence for the use of phytoestrogens as a
replacement for traditional estrogen replacement
therapy. Arch. Intern. Med. 161:1161–1172.
4. Setchell, K.D. 1998. Phytoestrogens: the biochemis-
try, physiology, and implications for human health
of soy isoflavones. Am. J. Clin. Nutr. 68:1333S–1346S.
5. Squadrito, F., et al. 2003. Effect of genistein on
endothelial function in postmenopausal women: a
randomized, double-blind, controlled study. Am. J.
6. Sacks, F.M. 2005. Dietary phytoestrogens to pre-
vent cardiovascular disease: early promise unful-
filled. Circulation. 111:385–387.
7. Cornwell, T., Cohick, W., and Raskin, I. 2004.
Dietary phytoestrogens and health. Phytochemistry.
8. Doerge, D.R., and Chang, H.C. 2002. Inactivation
of thyroid peroxidase by soy isoflavones, in vitro
and in vivo. J. Chromatogr. B Analyt. Technol. Biomed.
Life Sci. 777:269–279.
9. McVey, M.J., Cooke, G.M., and Curran, I.H. 2004.
Increased serum and testicular androgen levels in
F1 rats with lifetime exposure to soy isoflavones.
Reprod. Toxicol. 18:677–685.
10. Barnes, S. 2004. Soy isoflavones–phytoestrogens
and what else? J. Nutr. 134:1225S–1228S.
11. Mezei, O., et al. 2003. Soy isoflavones exert antidia-
betic and hypolipidemic effects through the PPAR
pathways in obese Zucker rats and murine RAW
264.7 cells. J. Nutr. 133:1238–1243.
12. Kim, S., et al. 2004. Genistein enhances expression
of genes involved in fatty acid catabolism through
activation of PPARalpha. Mol. Cell. Endocrinol.
13. Sack, M.N., et al. 1996. Fatty acid oxidation enzyme
gene expression is downregulated in the failing
heart. Circulation. 94:2837–2842.
14. Karbowska, J., Kochan, Z., and Smolenski, R.T.
2003. Peroxisome proliferator-activated receptor
alpha is downregulated in the failing human heart.
Cell. Mol. Biol. Lett. 8:49–53.
15. van der Schouw, Y.T., et al. 2005. Prospective study
on usual dietary phytoestrogen intake and cardio-
vascular disease risk in western women. Circulation.
16. Aronow, W.S., Ahn, C., and Kronzon, I. 1999. Com-
parison of incidences of congestive heart failure in
older African-Americans, Hispanics, and whites.
Am. J. Cardiol. 84:611–612, A619.
17. Aurigemma, G.P., and Gaasch, W.H. 1995. Gender
differences in older patients with pressure-over-
load hypertrophy of the left ventricle. Cardiology.
18. Krumholz, H.M., Larson, M., and Levy, D. 1993.
Sex differences in cardiac adaptation to isolated
systolic hypertension. Am. J. Cardiol. 72:310–313.
19. Greaves, S.C., Roche, A.H., Neutze, J.M., Whitlock,
R.M., and Veale, A.M. 1987. Inheritance of hyper-
trophic cardiomyopathy: a cross sectional and M
mode echocardiographic study of 50 families. Br.
Heart J. 58:259–266.
20. Olivetti, G., et al. 1995. Gender differences and
aging: effects on the human heart. J. Am. Coll. Car-
21. Pelzer, T., Shamim, A., Wolfges, S., Schumann, M.,
and Neyses, L. 1997. Modulation of cardiac hyper-
trophy by estrogens. In Hypertension and the heart.
Zanchetti, A., editor. Plenum Press. New York, New
York, USA. 83–89.
22. Ho, K.K., Anderson, K.M., Kannel, W.B., Grossman,
W., and Levy, D. 1993. Survival after the onset of
congestive heart failure in Framingham Heart
Study subjects. Circulation. 88:107–115.
23. Levy, D., et al. 2002. Long-term trends in the inci-
dence of and survival with heart failure. N. Engl. J.
24. Anderson, G.L., et al. 2004. Effects of conjugated
equine estrogen in postmenopausal women with
hysterectomy: the Women’s Health Initiative ran-
domized controlled trial. JAMA. 291:1701–1712.
25. Rossouw, J.E., et al. 2002. Risks and benefits of
estrogen plus progestin in healthy postmenopausal
women: principal results from the Women’s Health
Initiative randomized controlled trial. JAMA.
26. Marsh, J.D., et al. 1998. Androgen receptors medi-
ate hypertrophy in cardiac myocytes. Circulation.
27. Grohe, C., et al. 1997. Cardiac myocytes and fibro-
blasts contain functional estrogen receptors. FEBS
28. Olsson, M.C., Palmer, B.M., Leinwand, L.A., and
Moore, R.L. 2001. Gender and aging in a transgenic
mouse model of hypertrophic cardiomyopathy. Am.
J. Physiol. Heart Circ. Physiol. 280:H1136–H1144.
29. Vikstrom, K.L., Factor, S.M., and Leinwand, L.A.
1996. Mice expressing mutant myosin heavy chains
are a model for familial hypertrophic cardiomyopa-
thy. Mol. Med. 2:556–567.
30. Thigpen, J.E., et al. 1999. Phytoestrogen content
of purified, open- and closed-formula laboratory
animal diets. Lab. Anim. Sci. 49:530–536.
31. Thigpen, J.E., Setchell, K.D., Goelz, M.F., and Forsythe,
D.B. 1999. The phytoestrogen content of rodent diets.
Environ. Health Perspect. 107:A182–A183.
32. Johnson, B.D., et al. 1997. Increased expression of the
cardiac L-type calcium channel in estrogen receptor-
deficient mice. J. Gen. Physiol. 110:135–140.
33. Xu, Y., Arenas, I.A., Armstrong, S.J., and Davidge,
S.T. 2003. Estrogen modulation of left ventricu-
lar remodeling in the aged heart. Cardiovasc. Res.
34. Jalil, J.E., et al. 1989. Fibrillar collagen and myocar-
dial stiffness in the intact hypertrophied rat left
ventricle. Circ. Res. 64:1041–1050.
35. Hess, O.M., et al. 1981. Diastolic function and
myocardial structure in patients with myocardial
hypertrophy. Special reference to normalized visco-
elastic data. Circulation. 63:360–371.
36. Kawai, S., Okada, R., Kitamura, K., Suzuki, A., and
Saito, S. 1984. A morphometrical study of myocar-
dial disarray associated with right ventricular out-
flow tract obstruction. Jpn. Circ. J. 48:445–456.
37. Swynghedauw, B. 1986. Developmental and func-
tional adaptation of contractile proteins in cardiac
and skeletal muscles. Physiol. Rev. 66:710–771.
38. Tardiff, J.C., et al. 2000. Expression of the beta
(slow)-isoform of MHC in the adult mouse heart
causes dominant-negative functional effects. Am. J.
Physiol. Heart Circ. Physiol. 278:H412–H419.
39. Herron, T.J., Korte, F.S., and McDonald, K.S. 2001.
Loaded shortening and power output in cardiac
myocytes are dependent on myosin heavy chain
isoform expression. Am. J. Physiol. Heart Circ. Physiol.
40. Florini, J.R., Ewton, D.Z., and Foster, J.A. 1991.
Modern concepts of insulin-like growth factors. Elsevier.
New York, New York, USA. 487–503.
41. Decker, R.S., Cook, M.G., Behnke-Barclay, M., and
216? The?Journal?of?Clinical?Investigation http://www.jci.org Volume 116 Number 1 January 2006
Decker, M.L. 1995. Some growth factors stimulate
cultured adult rabbit ventricular myocyte hypertro-
phy in the absence of mechanical loading. Circ. Res.
42. Camper-Kirby, D., et al. 2001. Myocardial Akt acti-
vation and gender: increased nuclear activity in
females versus males. Circ. Res. 88:1020–1027.
43. Matsui, T., et al. 2002. Phenotypic spectrum caused
by transgenic overexpression of activated Akt in the
heart. J. Biol. Chem. 277:22896–22901.
44. Antos, C.L., et al. 2002. Activated glycogen syn-
thase-3 beta suppresses cardiac hypertrophy in
vivo. Proc. Natl. Acad. Sci. U. S. A. 99:907–912.
45. Klotz, D.M., et al. 2002. Requirement of estrogen
receptor-alpha in insulin-like growth factor-1 (IGF-1)-
induced uterine responses and in vivo evidence for
IGF-1/estrogen receptor cross-talk. J. Biol. Chem.
46. Martin, M.B., et al. 2000. A role for Akt in mediat-
ing the estrogenic functions of epidermal growth
factor and insulin-like growth factor I. Endocrinol-
47. Haq, S., et al. 2000. Glycogen synthase kinase-3beta
is a negative regulator of cardiomyocyte hypertro-
phy. J. Cell Biol. 151:117–130.
48. Choi, E.J., and Lee, B.H. 2004. Evidence for genis-
tein mediated cytotoxicity and apoptosis in rat
brain. Life Sci. 75:499–509.
49. Kumi-Diaka, J., Sanderson, N.A., and Hall, A.
2000. The mediating role of caspase-3 protease in
the intracellular mechanism of genistein-induced
apoptosis in human prostatic carcinoma cell lines,
DU145 and LNCaP. Biol. Cell. 92:595–604.
50. Thornberry, N.A., and Lazebnik, Y. 1998. Caspases:
enemies within. Science. 281:1312–1316.
51. Nagel, S.C., vom Saal, F.S., and Welshons, W.V. 1998.
The effective free fraction of estradiol and xeno-
estrogens in human serum measured by whole cell
uptake assays: physiology of delivery modifies estro-
genic activity. Proc. Soc. Exp. Biol. Med. 217:300–309.
52. Sharkey, L.C., et al. 1998. Effect of ovariectomy in
heart failure-prone SHHF/Mcc-facp rats. Am. J.
53. Sharkey, L.C., et al. 1999. Effect of ovariectomy and
estrogen replacement on cardiovascular disease in
heart failure-prone SHHF/Mcc- fa cp rats. J. Mol.
Cell. Cardiol. 31:1527–1537.
54. Schaible, T.F., Malhotra, A., Ciambrone, G., and
Scheuer, J. 1984. The effects of gonadectomy on left
ventricular function and cardiac contractile pro-
teins in male and female rats. Circ. Res. 54:38–49.
55. Gao, X.M., et al. 2003. Sex hormones and cardio-
myopathic phenotype induced by cardiac beta 2-
adrenergic receptor overexpression. Endocrinology.
56. Malhotra, A., Buttrick, P., and Scheuer, J. 1990.
Effects of sex hormones on development of physio-
logical and pathological cardiac hypertrophy in male
and female rats. Am. J. Physiol. 259:H866–H871.
57. Barkley, M.S., Lasley, B.L., Thompson, M.A., and
Shackleton, C.H. 1985. Equol: a contributor to
enigmatic immunoassay measurements of estro-
gen. Steroids. 46:587–608.
58. Butcher, R.L., Collins, W.E., and Fugo, N.W. 1974.
Plasma concentration of LH, FSH, prolactin,
progesterone and estradiol-17beta throughout
the 4-day estrous cycle of the rat. Endocrinology.
59. Nequin, L.G., Alvarez, J., and Schwartz, N.B. 1979.
Measurement of serum steroid and gonadotropin
levels and uterine and ovarian variables through-
out 4 day and 5 day estrous cycles in the rat. Biol.
60. Loeb, W.F., and Quimby, F.W. 1999. The clinical chemis-
try of laboratory animals. 2nd edition. Taylor & Francis
Group. Philadelphia, Pennsylvania, USA. 753 pp.
61. Brown, N.M., and Setchell, K.D. 2001. Animal
models impacted by phytoestrogens in commer-
cial chow: implications for pathways influenced by
hormones. Lab. Invest. 81:735–747.
62. Boettger-Tong, H., et al. 1998. A case of a labora-
tory animal feed with high estrogenic activity and
its impact on in vivo responses to exogenously
administered estrogens. Environ. Health Perspect.
63. Ratna, W.N. 2002. Inhibition of estrogenic stimu-
lation of gene expression by genistein. Life Sci.
64. Morkin, E. 2000. Control of cardiac myosin heavy
chain gene expression. Microsc. Res. Tech. 50:522–531.
65. [Anonymous]. 1977. Report of the American Insti-
tute of Nutrition ad hoc Committee on Standards
for Nutritional Studies. J. Nutr. 107:1340–1348.
66. Freeman, K., et al. 2001. Progression from hyper-
trophic to dilated cardiomyopathy in mice that
express a mutant myosin transgene. Am. J. Physiol.
Heart Circ. Physiol. 280:H151–H159.
67. Olsson, M.C., Palmer, B.M., Stauffer, B.L., Lein-
wand, L.A., and Moore, R.L. 2004. Morphological
and functional alterations in ventricular myocytes
from male transgenic mice with hypertrophic car-
diomyopathy. Circ. Res. 94:201–207.