s
ss
n
a
,
the uterus and the progression of the estrous cycle.
We propose that the liver acts as critical mediator
GnRH neurons are also the recipient of input from the neural
cells responsive to peripheral (e.g., leptin, insulin, ghrelin) (Magniet al., 2006; Lee et al., 2008); in most vertebrates, it is recognized
that malnutrition affects the reproductive cycle, but the molec-
ular events involved in this phenomenon remain to be clarified.
both ERa and ERb were shown to be instrumental in estrogen
regulation of the glucose transporter GLUT4 (Barros et al.,
2006); in the pancreas insulin content is regulated via ERaIn mammals the ovarian sex hormone 17b-estradiol (E2) and its
receptors ERa and ERb have a major role in the control of
reproductive functions as well as in the regulation of food intake
(Alonso-Magdalena et al., 2008). In the liver, where most of the
estrogen receptors are of the ERa type (Alvaro et al., 2000),
this receptor and estrogen were shown to control glucoseof energetic and reproductive functions responsible
for the blockade of the estrous cycle in case of
protein scarcity. Our findings may provide novel
insights to understand the cause of selected forms
of infertility and metabolic alterations in women after
menopause.
INTRODUCTION
Food intake and fertility are under a strict reciprocal control and
this mechanism ensures that reproduction occurs only in favor-
able conditions with respect to energy availability. The biochem-
ical pathways coupling nutrition and reproduction were
described in detail in invertebrates (Gerisch et al., 2001; Motola
et al., 1999; Chan and Mantzoros, 2001; Malik et al., 2008) and
central (NPY, orexins, kisspeptin, melanocortin) metabolic
hormones (Li et al., 1999; Campbell et al., 2003; Dungan et al.,
2006; Lee and Wardlaw, 2007). Thus the finding that ERs are
largely present in the hypothalamic regions which are respon-
sible for energy intake and storage (Musatov et al., 2007) and
that circulating levels of steroid hormones influence neurons
responsive to central metabolic hormones led to the formulation
of the hypothesis of a central mechanism for the control of ener-
getic metabolism by sex hormones (Schneider, 2004).
However, estrogen signaling is known to control some of the
mechanisms regulating the energetic metabolism in the periph-
eral organs as well. E2 controls the amount of white adipose
tissue (WAT) in female mice (Heine et al., 2000) and modulates
leptin production positively and negatively depending upon its
binding to ERa or ERb (Yi et al., 2008); in the skeletal muscle,the correct proliferation of the lumen epithelium in interactions (Herbison, 1998).Cell Metabolism
Article
Amino Acid-Dependent Act
Estrogen Receptor Alpha In
and Reproductive Function
Sara Della Torre,1,2 Gianpaolo Rando,1,2 Clara Meda,1,2 Ale
Cristian Ibarra,1,2 Paolo Magni,1,3 Paolo Ciana,1,2 and Adria
1Center of Excellence on Neurodegenerative Diseases
2Department of Pharmacological Sciences
3Department of Endocrinology, Pathophysiology and Applied Biology
University of Milan, Milan, 20133, Italy
4Institut de Genetique et de Biologie Moleculaire et Cellulaire Centre N
la Recherche Me´dicale, College de France, 67404 Illkirch-Strasbourg
*Correspondence: adriana.maggi@unimi.it
DOI 10.1016/j.cmet.2011.01.002
SUMMARY
Throughout evolution, organisms have devised strat-
egies to limit fertility in case of prolonged starvation.
In mammals, the liver plays a central role in the
orchestration of mechanisms allowing for the
maintenance of energy homeostasis. We here
demonstrate that dietary amino acids regulate the
transcriptional activity of hepatic estrogen receptor
alpha (ERa) through an mTOR-dependent mecha-
nism. As a result of ERa activation, hepatic IGF-1
mRNA and blood IGF-1 are increased. Conversely,
calorie restriction or selective ablation of ERa in the
liver decrease blood IGF-1 to levels inadequate forCell Mivation of Liver
tegrates Metabolic
via IGF-1
ia Stell,1,2 Pierre Chambon,4 Andre´e Krust,4
a Maggi1,2,*
tional de la Recherche Scientifique, Institut National de la Sante´ et de
France
(Dubuc, 1985; Asarian and Geary, 2006) and lipid and glucose
homeostasis (Hewitt et al., 2004); thus E2 signaling apparatus
is a potential candidate for the coupling of reproductive functions
with the energetic metabolism.
E2 is synthesized via a feedback mechanism involving the
hypothalamus in which the gonadotropin-releasing hormone
(GnRH) is produced to stimulate the anterior hypophysis to
release the gonadotropins into the bloodstream (Levine, 1997).
Gonadotropins elicit the maturation of follicles in the ovary and
the production of sex hormones that orchestrate the maturation
of the oocytes and the implantation of the fertilized eggs in the
uterus. GnRH synthesis in the hypothalamus is in turn controlled
by circulating E2 via complex mechanisms, including the
hormonal direct action on electrical activity of GnRH neurons
and indirect control exerted via neuron-neuron or glia-GnRHetabolism 13, 205–214, February 2, 2011 ª2011 Elsevier Inc. 205

acids exerted an effect selective for the sex hormone receptor
activity.
Because it is well known that sugars are rapidly catabolized,
we investigated the effects of C or AA administration at shorter
times after oral treatment. We measured luciferase activity in
living mice by BLI at 1, 3, 4, 7, and 8 hr after C or AA administra-
tion by gavage. At no time did C increase luciferase-dependent
photon emission in liver; conversely, 6–8 hr after AA administra-
tion photon emission was increased 4-fold versus time 0 (Fig-
ure 1F). We calculated that the total amount of AA we could
administer by gavage was equivalent to the 20% average daily
protein intake: considering that this amount of AA induced an
Figure 1. Food Consumption Is Associated with ER Transcriptional
Activity in the Liver
(A–D) Luciferase (LUC) activity as measured in vivo by BLI (A) and ex vivo by
enzymatic activity (RLU, B), LUC (C), and ERa mRNA (D) in liver prior to (0)
or at the end of weeks 1, 2, 3, and 4 of CR. Data represent the mean ± SEM
(n = 6). *p < 0.05; **p < 0.01 (versus time 0).
(E) LUC enzymatic activity in liver of mice 6 hr after gavage with isocaloric
amounts of carbohydrates (C), amino acids (AA), or lipids (L). Data represent
the mean ± SEM (n = 18).
(F) Time course of the effect of treatment with C or AA on liver LUC activity.
(G–I) BLI (G), LUC enzymatic activity (H), and LUC mRNA content (I) measured
in liver of mice 6 hr after gavage with AA and with or without ICI 182,780 (ICI).
Data represent the mean ± SEM (n = 6). *p < 0.05; **p < 0.01.homeostasis, improving glucose tolerance and insulin sensitivity
in animal models and humans (Takeda et al., 2003; Simpson
et al., 2005; Gao et al., 2006), and by regulating the activity of
specific genes, estrogen was described as directly stimulating
lipogenesis in several animal species (Courtney et al., 1988)
and the expression of lipogenic genes (Gowri et al., 2007). By
using the ERE-Luc reporter mouse model (Ciana et al., 2003a),
we clearly demonstrated amajor impact of E2 on liver ER activity;
most interestingly, liver ERa activity was strongly regulated by
food intake (Ciana et. al 2005) and during the female estrous
cycle ERa transcriptional activity in the liver and in the reproduc-
tive organs were temporally associated, in contrast to what was
observed in the nonreproductive organs (Ciana et al., 2003a).
These observations led us to further investigate the biological
relevance of food intake-dependent regulation of ER activity in
liver and to verify the extent to which liver ER activity could be
involved in the control of reproductive functions.
We here demonstrate the existence of a novel liver-dependent
mechanism controlling growth of the uterus and therefore
ensuring that regular ovulatory cycles occur only in relation to
an adequate nutritional supply.
RESULTS
Amino Acids Induce ER Transcriptional Activity in Liver
Previous studies with the ERE-Luc reporter mouse model
demonstrated that consumption of nonestrogenic food acti-
vated estrogen receptors (ER) in liver. To further demonstrate
the association between food consumption and liver ER activity,
we investigated the consequences of calorie restriction (CR).
Adult female ERE-Luc mice were subjected to 40% CR for
4 weeks and, at the end of each week, liver ER activity wasmoni-
tored in vivo by bioluminescence-based imaging (BLI, Figure 1A)
or ex vivo by quantitative analysis of luciferase enzymatic activity
(Figure 1B) and mRNA (Figure 1C). All measurements were done
in the morning (between 9 and 11 a.m.) when luciferase activity
was generally heightened by food consumption at night. CR
was associated with a very significant decrease in luciferase
mRNA and activity in the liver of ERE-Luc mice. This effect was
not associated with changes in ERa mRNA, indicating that the
receptor activity, but not its turnover, was affected by CR
(Figure 1D).
To identify the nature of the macronutrients responsible for
liver ER activity, we measured luciferase accumulation after
oral administration of an estrogen-free diet or an isocaloric
amount of its carbohydrate (C), protein (mainly amino acids), or
lipid (L) components. Figure 1E shows that 6 hr after gavage,
the complete diet induced an increase in luciferase activity of
67% versus controls; when each single macronutrient was
administered, the amino acids, but not the L or C, were able to
significantly increase ER activity (+46% versus control). As
control, we measured the effect of the diets on another well-
known sensor of cellular energy homeostasis: AMP-dependent
protein kinase (AMPK) (Andersson et al., 2004). As previously re-
ported, AMPK protein content in liver increased in animals fed
with the regular diet (+83%) and with C (+100%), while its phos-
phorylated state (PAMPK) augmented with C (+40%) and
decreased with the complete diet (
34%), amino acids
(
36%), and L (
40%) (Figure S1A, available online). Thus amino
206 Cell Metabolism 13, 205–214, February 2, 2011 ª2011 Elsevier InCell Metabolism
Liver Estrogen Receptor and Fertilityincrease in liver photon emission of the same order of magnitude
as endogenous estrogen at proestrus (P, the phase of the
estrous cycle in which circulating E2 is highest) (Figure S1B),
we concluded that liver ER activity induced by amino acids
may be associated with a specific physiological function. Thus
we further tested whether the AA effect on liver luciferase was
c.