Caveolin-1 Is Necessary for Hepatic Oxidative Lipid Metabolism: Evidence for Crosstalk between Caveolin-1 and Bile Acid Signaling

Abstract

Caveolae and caveolin-1 (CAV1) have been linked to several cellular functions. However, a model explaining their roles in mammalian tissues in vivo is lacking. Unbiased expression profiling in several tissues and cell types identified lipid metabolism as the main target affected by CAV1 deficiency. CAV1-/- mice exhibited impaired hepatic peroxisome proliferator-activated receptor α (PPARα)-dependent oxidative fatty acid metabolism and ketogenesis. Similar results were recapitulated in CAV1-deficient AML12 hepatocytes, suggesting at least a partial cell-autonomous role of hepatocyte CAV1 in metabolic adaptation to fasting. Finally, our experiments suggest that the hepatic phenotypes observed in CAV1-/- mice involve impaired PPARα ligand signaling and attenuated bile acid and FXRα signaling. These results demonstrate the significance of CAV1 in (1) hepatic lipid homeostasis and (2) nuclear hormone receptor (PPARα, FXRα, and SHP) and bile acid signaling.

Full text

Cell Reports
Report
Caveolin-1 Is Necessary for Hepatic Oxidative
Lipid Metabolism: Evidence for Crosstalk
between Caveolin-1 and Bile Acid Signaling
Manuel A. Ferna
´
ndez-Rojo,
1,9
Milena Gongora,
1
Rebecca L. Fitzsimmons,
1
Nick Martel,
1
Sheree D. Martin,
4
Susan J. Nixon,
1
Andrew J. Brooks,
1
Maria P. Ikonomopoulou,
1
Sally Martin,
1
Harriet P. Lo,
1
Stephen A. Myers,
1
Christina Restall,
5
Charles Ferguson,
1,2
Paul F. Pilch,
8
Sean L. McGee,
4
Robin L. Anderson,
5,6
Michael J. Waters,
1
John F. Hancock,
7
Sean M. Grimmond,
1
George E.O. Muscat,
1,3,
*
and Robert G. Parton
1,2,
*
1
Institute for Molecular Bioscience, The University of Queensland, Brisbane, Queensland 4067, Australia
2
Centre for Microscopy and Microanalysis, The University of Queensland, Brisbane, Queensland 4067, Australia
3
Obesity Research Centre, The University of Queensland, Brisbane, Queensland 4067, Australia
4
Metabolic Research Unit, School of Medicine, Deakin University, Geelong, Victoria 3217, Australia
5
Peter MacCallum Cancer Centre, Melbourne, Victoria 8006, Australia
6
Department of Pathology, The University of Melbourne, Melbourne, Victoria 8006, Australia
7
Department of Integrative Biology and Pharmacology, University of Texas Health Science Center at Houston, Houston, Texas 77030, USA
8
Department of Biochemistry, Boston University School of Medicine, Boston, MA 02118, USA
9
Present address: Monash University, Department of Biochemistry and Molecular Biology School of Biomedical Sciences Building 77,
Victoria 3800, Australia
*Correspondence: g.muscat@imb.uq.edu.au (G.E.O.M.), r.parton@imb.uq.edu.au (R.G.P.)
http://dx.doi.org/10.1016/j.celrep.2013.06.017
This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative Works
License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are
credited.
SUMMARY
Caveolae and caveolin-1 (CAV1) have been linked to
several cellular functions. However, a model explain-
ing their roles in mammalian tissues in vivo is lack-
ing. Unbiased expression profil ing in several tissues
and cell types identified lipid metabolism as the main
target affected by CAV1 deficiency. CAV1!/! mice
exhibited impaired hepatic peroxisome prolife rator-
activated receptor a (PPARa)-dependent oxidative
fatty acid metabolism and ketogenesis. Similar
results were recapitulated in CAV1-deficient AML12
hepatocytes, suggesting at least a partial cell-auton-
omous role of hepatocyte CAV1 in metabolic adap-
tation to fasting. Finally, our experiments suggest
that the hepatic phenotypes observed in CAV1!/!
mice involve impaired PPARa ligand signaling and
attenuated bile acid and FXRa signaling. Thes e
results demonstrat e the significance of CAV1 in (1)
hepatic lipid homeostasis and (2) nucl ear hormone
receptor (PPARa, FXRa, and SHP) and bile acid
signaling.
INTRODUCTION
Caveolae are plasma membrane microdomains enriched in
cholesterol and sphingolipids. They are extremely abundant in
specific cell types, including endothelial cells and adipocytes,
but not in cells such as hepatocytes (Calvo et al., 2001; Pilch
and Liu, 2011). Caveolin-1 (CAV1) is an integral membrane pro-
tein and, in association with PTRF/Cavin1, is the main structural
protein of caveolae in nonmuscle cells. Although caveolae and
CAV1 have been implicated in lipid regulation, mechanosensa-
tion, signaling, and endocytosis (Drab et al., 2001; Garg and
Agarwal, 2008; Hayashi et al., 2009; Liu et al., 2008; Parton
and Simons, 2007; Razani et al., 2002), a universal model that
explains the specific roles of caveolae and CAV1 in metabolism
in vivo is lacking.
Liver lipid metabolism can be physiologically challenged dur-
ing fasting and obesity when it becomes essential for systemic
energy homeostasis (Desvergne et al., 2006). In this context,
peroxisome proliferator-activated receptor a (PPARa) regulates
fatty acid oxidation (FAO) and ketogenesis. During fasting,
hepatic PPARa-dependent lipid metabolism depends on fatty
acid synthase (FAS) and diet-derived lipid metabolites that
work as endogenous PPARa ligands (Chakravarthy et al.,
2009; Chakravarthy et al., 2005). Moreover, bile acids (BAs)
and BA activation of FXRa are direct regulators of PPARa
signaling (Pineda Torra et al., 2003). The excessive accumulation
of BAs also provides protection against liver steatosis, but it
impairs FAO in mice (Watanabe et al., 2004). It has been pro-
posed that CAV1 is important in the maintenance of hepatic lipid
homeostasis (Ferna
´
ndez et al., 2006; Ferna
´
ndez-Rojo et al.,
2012; Martin and Parton, 2006; Parton and Simons, 2007).
Recently, several reports have described the involvement of
CAV1 in mitochondrial regulation (Asterholm et al., 2012; Bosch
et al., 2011). Asterholm et al. (2012) postulated that the metabolic
phenotype of tissues from CAV1!/! mice, including the liver, is
mainly caused by the adipocyte-CAV1 deficiency. However, the
lack of data showing the rescue of wild-type (WT) phenotypes by
238 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

the re-expression of CAV1 in adipocytes from CAV1!/! mice
means that some of the phenotypes observed in tissues from
CAV1!/! mice, including the liver, are totally or partially caused
by cell-specific CAV1 deficiency. Therefore, the mechanisms by
which CAV1 modulates hepatic lipid metabolism are still poorly
understood.
In this study, we used a wide range of approaches,
including microarray analysis, electron microscopy, bio-
chemistry, indirect calorimetry, three different CAV1!/!
mouse models, and two different CAV1-deficient AML12
hepatocyte cell models to show (1) that CAV1 participates in
multiple facets of lipid metabolism that are essential for the
maintenance of lipid homeostasis, (2) the universal and cell-
autonomous role of CAV1 in maintaining hepatocyte PPARa
signaling during fasting, and (3) CAV1-dependent BA and
FXRa signaling.
RESULTS
Lipid Metabolism Is the Major Cell and Tissue Process
Modulated by CAV1
Comparative genome-wide expression analysis in mouse
embryonic fibroblast (MEFs) primary cultures, livers, and
visceral adipose tissue from CAV1!/! and CAV1+/+ mice
(Drab et al., 2001) provided an unbiased approach for
investigating pathways directly or indirectly affected by CAV1
deficiency. We observed statistically significant alterations in
195, 421, and 288 genes in CAV1!/! MEFs, livers, and adipose
tissue, respectively, representing different sets of genes. A
remarkably small number of modulated genes were common
to two tissues or to all three tissues (see Table S1). The group
of genes modulated in all three tissues, predicted to represent
key universal proteins linked to CAV1 expression, included
signaling proteins such as K-Ras, sorting nexin 5, the sumoyla-
tion enzyme Ube21, and liprinb1, a phosphatase scaffolding
protein.
Enrichment of differentially expressed genes in several path-
ways and functions by ingenuity pathways analysis (IPA)
revealed that the only common cellular process affected by
CAV1 deficiency in adipose tissue, livers, and MEFs was a down-
regulation of lipid metabolism (Figure S1), demonstrating the
importance of CAV1 for the maintenance of lipid homeostasis.
Interestingly, the highest gene enrichment in lipid metabolism,
both the function and the pathway, was observed in CAV1!/!
livers (Figure 1A).
CAV1 in Fasting Hepatic Lipid Metabolism
Genome-wide expression profiling and gene enrichment anal-
ysis suggested that CAV1 deficiency altered the expression of
genes involved in hepatic peroxisomal and mitochondrial FAO,
bile synthesis, ketogenesis, lipid droplet formation, minor fatty
acid, and amino acid metabolism (Figure 1A).
In the mouse liver, these metabolic pathways are under the
control of the nuclear receptor PPARa (Mandard et al., 2004).
The expression of hepatic PPARa and its target genes is induced
6–8 hr after food deprivation and peaks at approximately 12 hr of
starvation (Yang et al., 2006). At this time of fasting, the transcript
levels of hepatic PPARa and its target genes, Cpt1a, Cpt1b,
and
PGC-1a, and the protein levels of PPARa and PGC-1a were
significantly reduced in CAV1 !/! mice in comparison to
CAV1+/+ mice ( Figures 1B–1D). Furthermore, decreased
expression of hepatic Bdh1 (Figure 1E) and reduced plasma
levels of the ketone body b-hydroxybutyrate (BOH) (Figure 1F)
also suggest that CAV1 deficiency impaired hepatic ketogen-
esis. Moreover, CAV1!/! MEFs oxidized less palmitate than
CAV1+/+ MEFs (Figure 1G).
Defective Fatty Acid Oxidation and Ketogenesis
Is Independent of the Lipodystrophic Phenotype Seen
in CAV1!/! Mice
Because fatty acids function as endogenous ligands for the
activation of PPARa (Chakravarthy et al., 2009; Chakravarthy
et al., 2005; Jump, 2011; Schroeder et al., 2008), we investigated
whether lipodystrophy in CAV1!/! mice (Drab et al., 2001;
Razani et al., 2002) might reduce available endogenous
fatty-acid-derived PPARa ligands and be responsible for the
impairment of PPARa signaling and ketogenesis. Neither sup-
plementation with oleic acid (OA), one of the main fatty acid
components of triacylglycerols (TAG) stored in hepatocyte lipid
droplets (LDs) (Fujimoto et al., 2006), nor with arachidonic acid,
the main precursor for endogenous PPARa ligands (Jump,
2011), recovered plasma BOH levels when compared to
untreated CAV1!/! and CAV1+/+ mice (Figure S2A).
Moreover, high-fat diet (HFD)-fed CAV1!/! mice, which still
have available fatty acids for FAO and ketogenesis in the form
of high levels of plasma lipids and cytoplasmic LDs in hepato-
cytes (Asterholm et al., 2012; Ferna
´
ndez-Rojo et al., 2012),
also showed lower levels of hepatic PPARa messenger RNA
(mRNA) and the PPARa target genes Cpt1a and MCAD than
HFD-fed
CAV1+/+ mice (Figure S2B). HFD-fed CAV1!/! mice
also exhibited decreased expression of the ketogenesis-related
genes Bdh1 and FGF21 as compared to HFD-fed CAV1+/+ mice
(Figure S2C). Finally, chow-fed CAV1!/! mice showed reduced
plasma BOH in comparison to chow-fed CAV1+/+ mice, which
was unchanged by HFD in the latter (Figure S2D).
In addition, microarray analysis (Figure S2E) and reduced pro-
tein levels of PPARa and PGC1a in CAV1!/! adipose tissue
(Figure S2F), as well as reduced expression of PPARa, Cpt1b,
and MCAD in CAV1!/! white adipose tissue explants in
comparison to CAV1+/+ explants (Figure S2G), supported the
hypothesis that CAV1 deficiency also deregulated adipocyte
mitochondrial function, PPARa signaling, and the expression of
FAO-related genes, despite the availability of fatty acids. Hence,
these results suggest that, independent of fatty acid availability,
CAV1 deficiency impairs PPARa signaling in metabolic tissues
such as liver and white adipose tissue.
Cell-Autonomous Modulation of Hepatocyte Fatty Acid
Catabolism and Mitochondrial Metabolic Adaptation by
CAV1
We examined the cell-autonomous role of caveolae and CAV1
in metabolic adaptation to fasting and hepatic mitochondrial
function during carbohydrate and lipid metabolism. We devel-
oped stable CAV1 knockdown (CAV1-kd) (Ferna
´
ndez-Rojo
et al., 2012) and PTRF/Cavin1 knockdown (PTRF/Cavin1-kd)
AML12 hepatocyte cell lines (Figure 2A and the Extended
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors 239

(legend on next page)
240 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

Experimental Procedures). Mimicking physiological conditions
in mice, AML12 hepatocyte cell lines were maintained in fed-
like (25 mM glucose) or fasting-like (10 mM glucose/100 mM
OA) culture conditions for 24 hr before we examined mitochon-
drial metabolic adaptation, oxidative phosphorylation—also
referred as mitochondrial respiration and presented as oxygen
consumption ratio—and cellular glycolysis—presented by the
extracellular acidification ratio—in living cells by using an XF
extracellular metabolic flux analyzer. Under normal fed-like
conditions, CAV1 deficiency was associated with a trend
toward a compensatory increase in glycolytic flux (Figure 2C)
without consequences for the ratios of mitochondrial respira-
tion ratios (Figure 2D). In agreement with previous data
(Ferna
´
ndez-Rojo et al., 2012), these results suggested that
CAV1 deficiency favors anaerobic glycolysis over oxidative
phosphorylation because of carbohydrate catabolism. Switch-
ing to fasting-like conditions in WT AML12 hepatocytes was
associated with a general conservation of nutrients, which
was reflected by a trend to decreased nonoxidative (glycolysis)
(Figures 2B and 2C) and oxidative metabolism (Figures 2B and
2D). However, CAV1 deficiency resulted in metabolic deregu-
lation when switching to fasting-like conditions, highlighted
by an increase in oxidative metabolism (Figures 2B and 2D)
and higher glycolytic flux (Figures 2B and 2C). On the other
hand, PTRF/Cavin1-kd cells showed no metabolic adaptation
to fasting-like conditions, but their mitochondrial respiration
did not differ from WT cells (Figures 2B and 2D). These
results suggested that CAV1 deficiency impairs hepatocyte
metabolic adaptation to fasting-like conditions, but not mito-
chondrial respiration, and reduces hepatocyte mitochondrial
capacity for metabolizing fatty acids, which is compensated
by an increase in nonoxidative and oxidative carbohydrate
consumption.
Universal Role of CAV1 in Liver Oxidative Lipid
Metabolism during Fasting
The deleterious effects of CAV1 deficiency on liver regeneration
and higher ratios of systemic carbohydrate metabolism in fed
mice (Asterholm et al., 2012; Ferna
´
ndez et al., 2006; Ferna
´
n-
dez-Rojo et al., 2012; Mayoral et al., 2007) depend on the genetic
background. In contrast, by studying
Balb/C
CAV1!/! mice (Fer-
na
´
ndez-Rojo et al., 2012) and
JAX
CAV1!/! mice obtained from
the Jackson Laboratory (Razani et al., 2002), we observed
decreased hepatic PPARa protein levels, reduced MCAD and
PDK4 expression, and a low concentration of plasma BOH in
24 hr-fasted
Balb/C
CAV1!/! mice (Figures S3A–S3C). Similar
results were obtained in 24 hr-fasted
JAX
CAV1!/! mice, which
showed reduced MCAD and Bdh1 expression and a low concen-
tration of plasma BOH (Figures S3D and S3E). Furthermore,
similar results were obtained by comparing 24 hr-fasted PTRF/
Cavin1!/! and PTRF/Cavin1+/+ littermates. PTRF/Cavin1!/!
mice lack caveolae in all their tissues (Bastiani et al., 2009; Liu
et al., 2008) but expressed 14% and 18% of WT CAV1 levels in
fed and fasted PTRF/Cavin1!/! livers, respectively (Figure S3F).
Similar to the three CAV1!/! mouse strains, 24 hr-fasted PTRF/
Cavin1!/! mice showed reduced levels of total and phosphor-
ylated PPARa protein in liver homogenates (Figure S3G), corre-
lating with the defective expression of PPARa-dependent genes
involved in FAO and ketogenesis and lower levels of plasma BOH
(Figures S3H–S3J). Hence, our results demonstrated universal
and genetic-background-independent regulation of hepatic
oxidative lipid metabolism, including PPARa-dependent meta-
bolism and ketogenesis by CAV1.
The Absence of CAV1 Confers Resistance to Wy14643-
Induced Activation of PPARa Signaling during Fasting
To emphasize the significance of CAV1 in hepatic lipid meta-
bolism and metabolic diseases, we tested whether the specific
PPARa agonist wy14643 rescued PPARa-dependent path-
ways in fasted 9- to 12-month-old CAV1+/+ and CAV1!/!
mice (old
k
CAV1+/+ and
k
CAV1!/! mice). Comparatively, 9-
to 12-month-old mice are equivalent to humans that are 35–40
years old, the age at which most metabolic diseases appear in
humans (Curtis et al., 2005).
Similar to young mice, CAV1 deficiency reduced the expres-
sion of PPARa target genes and plasma ketone bodies in old
mice (Figures 3A–3D). In order to recover PPARa signaling, old
CAV1+/+ and CAV1!/! mice were starved for the last 24 hr of
the 7-day wy14643 treatment. Wy14643 treatment increased
total and active hepatic PPARa protein levels in old CAV1+/+,
but not in CAV1!/!, mice (Burns and Vanden Heuvel, 2007)(Fig-
ures 3B and 3C). Wy14643 treatment recovered the hepatic
expression of the PPARa target gene MCAD, but not Bdh1,in
24 hr-fasted CAV1!/! mice (Figure 3A). Accordingly, wy14643
did not stimulate ketogenesis in old CAV1!/! mice (Figure 3D).
Interestingly, plasma ketone body concentration in untreated old
CAV1!/! mice was significantly lower than it was in young
CAV1!/! mice (Figures 3D–3D1), suggesting that CAV1 defi-
ciency might confer more dramatic metabolic consequences
during aging. Similar to young mice (Ferna
´
ndez-Rojo et al.,
Figure 1. CAV1 Deficiency Reduces Hepatic FAO, Ketogenesis, and Systemic Energy Metabolism
(A) A heat plot representing decreased (shades of blue) and increased (shades of green) expression of genes involved in hepatic lipid and energy metabolism in
fasted CAV1!/! mice.
(B and C) Hepatic PPARa (B) and PGC1a (C) mRNA and protein in 9 hr- to 12 hr-fasted CAV1!/! mice.
(D and E) Hepatic Cpt1a, Cpt1b, and MCAD expression (D) and Bdh1 expression (E) in 9 hr- to 12 hr-fasted CAV1!/! (white bars) and CAV1+/+ (black bars) mice.
(F) Plasma b -hydroxybutyrate (BOH) levels in CAV1+/+ and CAV1!/! mice (n = 9).
(G) Oxidation of palmitate in CAV1!/! MEFs (n = 9).
(H and I) Respiratory exchange ratio (RER) (H) and volume of oxygen consumed (I) during fasting in CAV1+/+ (blue circles) and CAV1!/! (red squares) mice (n = 10
and 7, respectively).
(J) Amount of systemic fat oxidized (J) and systemic calories produced (K) during fasting by CAV1!/! (white bars) and CAV1+/+ (black bars) mice (n = 10 and 7,
respectively).
The data represent the mean ± SEM. In (B)–(G) and (J) and (K), the statistical significance was assessed with a Student’s t test, whereas, in (H) and (I), we used a
one-way ANOVA test. *p < 0.05, **p < 0.01, ***p < 0.001.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors 241

2012), old CAV1!/! also showed impaired fasting-induced
steatosis when compared to old CAV1+/+ mice (Figure 3F). In
accordance with the wy14643-induced overactivation of PPARa
signaling and FAO together with and increased mobilization of
LD-TAG, ADRP mRNA and protein levels, as well as the number
of LDs, were reduced in wy14643-treated old
k
CAV1+/+ mice.
However, neither ADRP mRNA and protein levels nor the
number of LDs were reduced in livers from wy14643-treated
old
k
CAV1!/! mice (Figures 3A, 3B, and 3F). Furthermore, small
interfering RNA (siRNA) against CAV1 in AML12 hepatocytes
inhibited the stimulation of PPARa and PPARa target genes in
response to wy14643 recapitulating the phenotype seen in
CAV1!/! livers (Figure 3E) and supporting the cell-autonomous
role of CAV1 in hepatocyte PPARa signaling. Intriguingly, and
unlike in fasted CAV1!/! mice, wy14643-stimulated PPAR a
targets genes in fed ad libitum CAV1!/! mice. Wy14643
induced MCAD and ADRP, and their expression levels were
similar to or higher than those seen in CAV1+/+ mice (Figure 3G),
whereas Bdh1 transcript levels were unaffected in either mouse
strains, probably because the activation of ketogenesis is
Figure 2. Cell-Autonomous Role of Hepatocyte CAV1 on Fasted-Associated Oxidative Lipid Metabolism and Mitochondrial Function
(A) CAV1 and PTRF /Cavin1 protein levels in wild-type (WT), CAV1 knockdown (CAV1-kd#2 and CAV1-#4), and PTRF/Cavin1 knockdown (PTRF-kd#3) AML12
hepatocytes.
(B) Cellular bioenergetics in WT, CAV1-kd, and PTRF/Cavin1-kd AML12 hepatocytes cultured in fed-like (HG) or in fasted-like (LG/OA) medium. ECAR, extra-
cellular acidification rate; OCR, oxygen consumption rate.
(C) Basal glycolysis.
(D) Basal mitochondrial respiration.
The data represent the mean ± SEM. Statistical significance was assessed with a one-way ANOVA test. For inter-cell-type analysis (asterisks) and interdiet
analysis (hash marks); */#p < 0.05, **/##p < 0.01, ***/###p < 0.001.
242 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

(legend on next page)
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors 243

unnecessary in light of the availability of glucose in the peripheral
tissues.
CAV1 Deficiency Impairs the BA-FXRa Signaling Axis
in Hepatocytes
BAs, through the activation of FXRa, are essential regulators of
PPARa (Pineda Torra et al., 2003). Also, BAs regulate CAV1
expression in esophageal epithelial cells (Prade et al., 2012).
Hence, we investigated the potential crosstalk between BA
signaling and CAV1 by examining the BA- and FXRa-dependent
induction of the expression of short heterodimer partner (SHP,a
nuclear receptor), which is a critical regulator of hepatic choles-
terol and BA homeostasis (Goodwin et al., 2000; Wang et al.,
2002). CAV1-kd AML12 hepatocytes showed a dramatic loss
of BA-induced SHP expression (Figure 4A) relative to BA-depen-
dent induction of SHP in WT cells. Moreover, the induction of
SHP expression in response to the synthetic FXRa agonist
GW4064 was also attenuated in CAV1-kd AML12 hepatocytes
relative to WT cells (Figure 4B). Consistent with experiments in
AML12 hepatocytes and in correlation with impaired hepatic
PPARa signaling, 24 hr-fasted CAV1!/! mice displayed
decreased hepatic SHP expression (Figure 4C), providing
evidence for physiological crosstalk between CAV1 and hepato-
cyte BA signaling in vivo. No difference in hepatic FXRa expres-
sion between CAV1!/! and CAV1+/+ mice was observed
(Figure S4A). However, further evidence supporting attenuated
BA signaling in vivo was gleaned from the hepatic expression
profiling analysis, which revealed significant upregulation of
the estrogen receptor a (Figure S4B), an SHP-repressed gene
(Seol et al., 1998).
Furthermore, in CAV1!/! mice, the expression of hepatic
cholesterol 7a hydroxylase (Cyp7a1), whose product catalyzes
the rate-limiting step in the synthesis of BAs, was significantly
higher than it was in CAV1+/+ mice (Figure 4D). In mice,
Cyp7a1 expression is mainly suppressed by the activation of in-
testinal FXRa,
enterocyte expression and secretion of FGF15,
and, to a lesser degree, the hepatic FXRa-SHP axis (Goodwin
et al., 2000; Kong et al., 2012). Hence, the deregulation of
Cyp7a1 will be indicative of a defective intestinal FXRa-FGF15
signaling axis.
Finally, the volume of the accumulated BA in the gallbladders
of fasted CAV1!/! mice was lower than that of CAV1+/+ mice
(Figure S4C), suggesting that CAV1 might also be important for
intracellular trafficking and secretion into the canaliculi of the
BAs. Interestingly, in comparison to fasted CAV1+/+ mice,
serum BA levels were also reduced in 24 hr-fasted CAV1!/!
mice (Figure S4D). In summary, our data suggest that CAV1
modulates BA signaling, synthesis, and trafficking.
DISCUSSION
The genome-wide expression profile shown in this study sup-
ports the notion that CAV1 participates in multiple facets of lipid
metabolism. Specifically, our data highlight the metabolic signif-
icance of the systemic expression of CAV1 for hepatic lipid
metabolism. CAV1 maintains liver oxidative lipid metabolism
and ketogenesis during fasting and high-fat feeding indepen-
dently of the genetic background and the availability of fatty
acids. Alternatively, our data suggest that CAV1 might control
oxidative lipid metabolism in other cell types, such as fibroblasts
and adipocytes.
Moreover, two different and independent CAV1-deficienct
AML12 hepatocyte models supported the cell-autonomous
role of hepatocyte CAV1 in hepatic metabolic adaptation to fast-
ing and energy homeostasis. This conclusion is consistent with
previous studies overexpressing CAV1 or Caveolin-3 in vivo in
rodent livers (Frank et al., 2001; Otsu et al., 2010). These data
are fundamental for understanding the implications of CAV1 in
liver and systemic mammalian lipid metabolism and energy ho-
meostasis. Furthermore, the data argue against recent models
that suggest that all the phenotypes observed in CAV1!/! livers
are due to adipocyte CAV1 deficiency (Asterholm et al., 2012).
Other studies have demonstrated that cholesterol accumulation
in the mitochondria underlies mitochondrial dysfunction (Bosch
et al., 2011). However, selective diet-dependent activation of
hepatic PPARa signaling in CAV1!/! mice by wy14643 treat-
ment argues against mitochondrial cholesterol overload being
responsible for the defective metabolic adaptation to starvation
in CAV1 !/! mice. In a similar manner, given that BA and FXRa
signaling operates independently of mitochondrial function, it
would be unlikely that mitochondrial cholesterol overloading
attenuates SHP expression.
Evidence showing that CAV1 deficiency impairs BA and FXRa
signaling, which is necessary for PPARa expression, suggested
crosstalk between CAV1 and nuclear receptor signaling (i.e.,
FXRa, SHP, and PPARa). Indeed, the hallmark of attenuated
BA signaling (i.e., significantly reduced SHP expression) and
elevated expression of liver Cyp7a1, whose regulation depends
on the activation of the intestinal FXRa-FGF15 signaling axis
(Kong
et al., 2012), suggested that liver lipid metabolism might
Figure 3. Defective Wy14643-Induced PPAR a Signaling in Fasted CAV1!/! Mice
(A) Gene expression of PPARa target genes MCAD, ADRP , and Bdh1 in 24 hr-fasted mice. Nontreated CAV1+/+ mice are represented by the red dot line.
(B and C) ADRP, PPARa, CAV1, and actin protein levels in liver homogenates and purified lipid droplet fractions. Total, inactive, and active PPARa protein signal
quantification was obtained with ImageJ.
(D) Plasma BOH levels. In (D.1), the graph shows fold differences in plasma BOH concentrations between young, old, and wy-14643-treated old CAV1! /! and
CAV1+/+ mice.
(E) Relative expression fold change of PPARa target genes in untreated (gray bar) and in wy14643 (10 mM, white bar)-treated siRNA CAV1-kd AML12 hepatocytes
in comparison to untreated and wy14643 (10 mM)-treated WT AML12 hepatocytes (both represented by the gray broken line), respectively, (n = 3).
(F) Electron microscopy in livers.
(G) Gene expression of PPARa-target genes MCAD, ADRP, and Bdh1 in 7 day wy14643 (100 mM)-treated, fed ad libitum CAV1+/+ and CAV1!/! mice.
In (A)–(D) and (F), the experiments carried out in 7 day wy14643 (100 mM)-treated, 24 hr-fasted CAV1+/+ and CAV1!/! mice, and nontreated CAV1!/! and
CAV1+/+ mice.
The data represent the mean ± SEM.
244 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

Figure 4. CAV1 Deficiency Impairs Hepatic BA and FXRa Signaling
(A and B) SHP expression in 40 mM BA-treated (A) and 3 mM GW404-treated (B) AML12 hepatocytes.
(C and D) Liver SHP (C) and Cyp7a1 (D) expression in 24 hr-fasted CAV1+/+ and CAV1!/! mice.
(E) Hypothetical model of the deregulation of PPARa signaling in CAV1!/! livers.
The data represent the mean ± SEM. The statistical significance was assessed with a Student’s t test. *p < 0.05, **p < 0.01, ***p < 0.001.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors 245

be regulated by CAV1-dependent modulation of nuclear recep-
tors in the liver and other tissues, such as the small intestine.
We hypothesize that the lack of CAV1 might modify the lipid
and protein organization and/or composition of the hepatocyte
plasma membrane, which is required for BA-dependent SHP
expression. In addition, our results show dramatically attenuated
GW4064-dependent activation of FXRa in CAV1-kd hepatocytes
showing additional downstream effects on this signaling
pathway. Furthermore, CAV1 reduced gallbladder bile acid vol-
ume in CAV1!/! mice, suggesting that CAV1 also regulates
BA secretion into the canaliculi. Despite low gallbladder bile
volume, circulating serum bile levels are not increased in fasted
CAV1!/! mice. Instead, they tend to be lower than those of
CAV1+/+ mice, suggesting that BA might accumulate in
hepatocytes.
On the basis of these data, we hypothesize that, when hepatic
lipid metabolism is challenged in CAV1!/! mice, such as during
fasting and HFD feeding, PPARa signaling is compromised
because CAV1 deficiency leads to the hepatic accumulation of
BAs (Figure 4E). BA accumulation, coupled with BA detergent
properties, is consistent with resistance to developed hepatos-
teatosis and defective PPARa signaling (Watanabe et al.,
2004). Despite the availability of FAS and diet ‘‘new fat’’-derived
liver PPARa ligands by fasted and HFD-fed CAV1!/! mice, BA
accumulation might perturb the storage and/or compartmen-
talization, trafficking, and/or function of endogenous PPARa
ligands and impair PPARa signaling. In a similar manner, hepato-
cyte BA accumulation in fasted CAV1!/! mice may explain fast-
ing-specific impaired function of the synthetic PPARa agonist
wy14643. These defects would be the combination of the dele-
terious effects that a lack of CAV1 has on (1) the regulation of
the hepatocyte endocytic network (Parton and Simons, 2007;
Woudenberg et al., 2010), (2) the mobilization of arachidonic
acid (Astudillo et al., 2011), and (3) BA and FXRa signaling. Inter-
estingly, wy14643-dependent regulation of hepatic expression
of ADRP in ad libitum fed, but not fasted, CAV1!/! mice sug-
gested that defective steatosis in CAV1!/! mice ( Asterholm
et al., 2012; Ferna
´
ndez et al., 2006; Ferna
´
ndez-Rojo et al.,
2012) might also be caused by the disruption of PPARa ligand
function in hepatocytes during metabolic challenges.
This study provides a framework for understanding how CAV1
modulates lipid metabolism in a tissue-specific manner and
underscores the significance of the genetic background for the
development of global and tissue-specific mouse models. More-
over, the significance of our study is underscored by the central
position of the liver in the regulation of lipid and glucose homeo-
stasis and emphasizes the relevance of hepatocyte CAV1 for
maintaining hepatic mitochondrial function in the context of liver
and systemic lipid homeostasis.
EXPERIMENTAL PROCEDURES
Animals and Reagents
K
CAV1+/+ and
K
CAV1!/! mice were obtained as described in Ferna
´
ndez et al.
(2006). Liver sample collection and
JAX
CAV1!/! mice and
Balb/C
CAV1!/! mice
were generated as described in Ferna
´
ndez-Rojo et al. (2012). PTRF/Cavin1!/!
mice were generated as described in Liu et al. (2008). For experiments in
Balb/C
CAV1!/! mice and PTRF/Cavin1!/! mice, we used their matching
CAV1+/+ littermates. Mice were kept under a controlled humidity and lighting
schedule with a 12 hr dark period. All animals received care in compliance
with institutional guidelines regulated by the Australian government. HFD
feeding was performed as described in Ferna
´
ndez-Rojo et al. (2012). For fasting
experiments, food withdrawal was initiated at 6 a.m. after animal house lights
were switched on. When applicable, CAV1!/! mice were provided with
500 ml of 4 mM of OA and arachidonic acid by intraperitoneal injection. Mice
10–18 weeks old were fasted for up to 24 hr prior to experimentation.
Wy14643 was obtained from Sigma-Aldrich (C7081). CAV1 antibody was
obtained from BD Biosciences (#610060), and ADRP was obtained from Pro-
gen Biotechnik (#GP40). PPARa (Cayman Chemycal), PGC1a (Santa Cruz
Biotechnology, H-300), PTRF/Cavin1 antibody as described in Bastiani et al.
(2009), and mouse b-actin antibody were from Chemicon (#MAB1501).
BAs and GW4064 FXRa Agonist Treatment of AML12 Hepatocytes
Prior to RNA purification, WT, CAV1-kd, and PTRF/Cavin1-kd AML12 hepato-
cytes were treated with 40 mM of cholic acid and chenodeoxycholic acid or
with 3 mM GW4064 in 10% heat inactivated serum supreme (HISS)-supple-
mented Dulbecco’s modified Eagle’s serum/F12 medium for 24 hr.
Bile Collection from Gallbladders
Prior to liver resection, gallbladder-stored bile from 12 hr- and 24 hr-fasted
CAV1+/+ and CAV1!/! mice was harvested with syringes with 30G needles.
We quantified the harvested bile using micropipettes.
ACCESSION NUMBERS
The microarray data for MEFs, adipose tissue, and liver tissue of CAV1!/!
mice have been deposited in the Gene Expression Omnibus under accession
number GSE19045.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Extended Experimental Procedures, four
figures, and one table and can be found with this article online at http://dx.
doi.org/10.1016/j.celrep.2013.06.017.
ACKNOWLEDGMENTS
This work was supported by grants from the National Health and Medical
Research Council (NHMRC) of Australia (to R.G.P., G.E.O.M., S.G., and
R.L.A.) and from the National Breast Cancer Foundation (R.L.A.). M.A.F.- R.
was supported by the Program of MEC/Fulbright postdoctoral fellowships
from the Spanish government, the Diabetes Australia Research Trust, and
the NHMRC. We thank all the members of Parton laboratory for support and
discussion. The authors acknowledge the facilities as well as scientific
and technical assistance from the staff in the Australian Cancer Research Foun-
dation, the Institute for Molecular Bioscience Dynamic Imaging Facility for
Cancer Biology, and the Australian Microscopy and Microanalysis Facility at
the Centre for Microscopy and Microanalysis at the University of Queensland.
Received: August 26, 2012
Revised: March 4, 2013
Accepted: June 14, 2013
Published: July 11, 2013
REFERENCES
Asterholm, I.W., Mundy, D.I., Weng, J., Anderson, R.G., and Scherer, P.E.
(2012). Altered mitochondrial function and metabolic inflexibility associated
with loss of caveolin-1. Cel l Metab. 15, 171–185.
Astudillo, A.M., Pe
´
rez-Chaco
´
n, G., Meana, C., Balgoma, D., Pol, A., Del Pozo,
M.A., Balboa, M.A., and Balsinde, J. (2011). Altered arachidonate distribution
in macrophages from caveolin-1 null mice leading to reduce d eicosanoid
synthesis. J. Biol. Chem. 286, 35299–35307.
Bastiani, M., Liu, L., Hill, M.M., Jedrychowski, M.P., Nixon, S.J., Lo, H.P.,
Abankwa, D., Luetterforst, R., Fernandez-Rojo, M., Breen, M.R., et al.
246 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

(2009). MURC/Cavin-4 and cavin family members form tissue-specific caveo-
lar complexes. J. Cell Biol. 185, 1259–1273.
Bosch, M., Marı
´,
M., Herms, A., Ferna
´
ndez, A., Fajardo, A., Kassan, A., Giralt,
A., Colell, A., Balgoma, D., Barbero, E., et al. (2011). Caveolin-1 deficiency
causes cholesterol-dependent mitochondrial dysfunction and apoptotic
susceptibility. Curr. Biol. 21, 681–686.
Burns, K.A., and Vanden Heuvel, J.P. (2007) . Modulation of PPAR activity via
phosphorylation. Biochim. Biophys. Acta 1771, 952–960.
Calvo, M., Tebar, F., Lopez-Iglesias, C., and Enrich, C. (2001). Morphologic
and functional characterization of caveolae in rat liver hepatocytes. Hepatol-
ogy 33, 1259–1269.
Chakravarthy, M.V., Pan, Z., Zhu, Y., Tordjman, K., Schneider, J.G., Coleman,
T., Turk, J., and Semenkovich, C.F. (2005). ‘‘New’’ hepatic fat activates
PPARalpha to maintain glucose, lipid, and cholesterol homeostasis. Cell
Metab. 1, 309–322.
Chakravarthy, M.V., Lodhi, I.J., Yin, L., Malapaka, R.R., Xu, H.E., Turk, J., and
Semenkovich, C.F. (2009). Identification of a physiologically relevant endoge-
nous ligand for PPARalpha in liver. Cell 138, 476–488.
Curtis, R., Geesaman, B.J., and DiStefano, P.S. (2005). Ageing and meta-
bolism: drug discovery opportunities. Nat. Rev. Drug Discov. 4, 569–580.
Desvergne, B., Michalik, L., and Wahli, W. (2006). Transcriptional regulation of
metabolism. Physiol. Rev. 86, 465–514.
Drab, M., Verkade, P., Elger, M., Kasper, M., Lohn, M., Lauterbach, B., Menne,
J., Lindschau, C., Mende, F., Luft, F.C., et al. (2001). Loss of caveolae, vascular
dysfunction, and pulmonary defects in caveolin-1 gene-disrupted mice.
Science 293, 2449–2452.
Ferna
´
ndez, M.A., Albor, C., Ingelmo-Torres, M., Nixon, S.J., Ferguson, C.,
Kurzchalia, T., Tebar, F., Enrich, C., Parton, R.G., and Pol, A. (2006). Caveo-
lin-1 is essential for liver regeneration. Science 313, 1628–1632.
Ferna
´
ndez-Rojo, M.A., Restall, C., Ferguson, C., Martel, N., Martin, S., Bosch,
M., Kassan, A., Leong, G.M., Martin, S.D., McGee, S.L., et al. (2012). Caveolin-1
orchestrates the balance between glucose and lipid-dependent energy meta-
bolism: implications for liver regeneration. Hepatology 55, 1574–1584.
Frank, P.G., Pedraza, A., Cohen, D.E., and Lisanti, M.P. (2001). Adenovirus-
mediated expression of caveolin-1 in mouse liver increases plasma high-den-
sity lipoprotein levels. Biochemistry 40, 10892–10900.
Fujimoto, Y., Onoduka, J., Homma, K.J., Yamaguchi, S., Mori, M., Higashi, Y.,
Makita, M., Kinoshita, T., Noda, J., Itabe, H., and Takanoa, T. (2006). Long-
chain fatty acids induce lipid droplet formation in a cultured human hepatocyte
in a manner dependent of Acyl-CoA synthetase. Biol. Pharm. Bull. 29, 2174–
2180.
Garg, A., and Agarwal, A.K. (2008). Caveolin-1: a new locus for human lipodys-
trophy. J. Clin. Endocrinol. Metab. 93, 1183–1185.
Goodwin, B., Jones, S.A., Price, R.R., Watson, M.A., McKee, D.D., Moore,
L.B., Galardi, C., Wilson, J.G., Lewis, M.C., Roth, M.E., et al. (2000). A regula-
tory cascade of the nuclear receptors FXR, SHP-1, and LRH-1 represses bile
acid biosynthesis. Mol. Cell 6, 517–526.
Hayashi, Y.K., Matsuda, C., Ogawa, M., Goto, K., Tominaga, K., Mitsuhashi,
S., Park, Y.E., Nonaka, I., Hino-Fukuyo, N., Haginoya, K., et al. (2009). Human
PTRF mutations cause secondary deficiency of caveolins resulting in muscular
dystrophy with generalized lipodystrophy. J. Clin. Invest. 119, 2623–2633.
Jump, D.B. (2011). Fatty acid regulation of hepatic lipid metabolism. Curr.
Opin. Clin. Nutr. Metab. Care 14, 115–120.
Kong, B., Wang, L., Chiang, J.Y., Zhang, Y., Klaassen, C.D., and Guo, G.L.
(2012). Mechanism of tissue-specific farnesoid X receptor in suppressing the
expression of genes in bile-acid synthesis in mice. Hepatology 56, 1034–1043.
Liu, L., Brown, D., McKee, M., Lebrasseur, N.K., Yang, D., Albrecht, K.H.,
Ravid, K., and Pilch, P.F. (2008). Deletion of Cavin/PTRF causes global loss
of caveolae, dyslipidemia, and glucose intolerance. Cell Metab. 8, 310–317.
Mandard, S., Mu
¨
ller, M., and Kersten, S. (2004). Peroxisome proliferator-acti-
vated receptor alpha target genes. Cell. Mol. Life Sci. 61, 393–416.
Martin, S., and Parton, R.G. (2006). Lipid droplets: a unified view of a dynamic
organelle. Nat. Rev. Mol. Cell Biol. 7, 373–378.
Mayoral, R., Ferna
´
ndez-Martı
´nez,
A., Roy, R., Bosca
´
,
L., and Martı
´n-Sanz,
P.
(2007). Dispensability and dynamics of caveolin-1 during liver regeneration
and in isolated hepatic cells. Hepatology 46, 813–822.
Otsu, K., Toya, Y., Oshikawa, J., Kurotani, R., Yazawa, T., Sato, M., Yo-
koyama, U., Umemura, S., Minamisawa, S., Okumura, S., and Ishikawa, Y.
(2010). Caveolin gene transfer improves glucose metabolism in diabetic
mice. Am. J. Physiol. Cel l Physiol. 298, C450–C456.
Parton, R.G., and Simons, K. (2007). The multiple faces of caveolae. Nat. Rev.
Mol. Cell Biol. 8, 185–194.
Pilch, P.F., and Liu, L. (2011). Fat caves: caveolae, lipid trafficking and lipid
metabolism in adipocytes. Trends Endocrinol. Metab. 22, 318–324.
Pineda Torra, I., Claudel, T., Duval, C., Kosykh, V., Fruchart, J.C., and Staels,
B. (2003). Bile acids induce the expression of the human peroxisome prolifer-
ator-activated receptor alpha gene via activation of the farnesoid X receptor.
Mol. Endocrinol. 17, 259–272.
Prade, E., Tobiasch, M., Hitkova, I., Scha
¨
ffer, I., Lian, F., Xing, X., Ta
¨
nzer, M.,
Rauser, S., Walch, A., Feith, M., et al. (2012). Bile acids down-regulate caveo-
lin-1 in esophageal epithelial cells through sterol responsive element-binding
protein. Mol. Endocrinol. 26, 819–832.
Razani, B., Combs, T.P., Wang, X.B., Frank, P.G., Park, D.S., Russell, R.G., Li,
M., Tang, B., Jelicks, L.A., Scherer, P.E., and Lisanti, M.P. (2002). Caveolin-1-
deficient mice are lean, resistant to diet-induced obesity, and show hypertri-
glyceridemia with adipocyte abnormalities. J. Biol. Chem. 277, 8635–8647.
Schroeder, F., Petrescu, A.D., Huang, H., Atshaves, B.P., McIntosh, A.L., Mar-
tin, G.G., Hostetler, H.A., Vespa, A., Landrock, D., Landrock, K.K., et al. (2008).
Role of fatty acid binding proteins and long chain fatty acids in modulating nu-
clear receptors and gene transcription. Lipids 43, 1–17.
Seol, W., Hanstein, B., Brown, M., and Moore, D.D. (1998). Inhibition of estro-
gen receptor action by the orphan receptor SHP (short heterodimer partner).
Mol. Endocrinol. 12, 1551–1557.
Wang, L., Lee, Y.K., Bundman, D., Han, Y., Thevananther, S., Kim, C.S., Chua,
S.S., Wei, P., Heyman, R.A., Karin, M., and Moore, D.D. (2002). Redundant
pathways for negative feedback regulation of bile acid production. Dev. Cell
2, 721–731.
Watanabe, M., Houten, S.M., Wang, L., Moschetta, A., Mangelsdorf, D.J.,
Heyman, R.A., Moore, D.D., and Auwerx, J. (2004). Bile acids lower triglyceride
levels via a pathway involving FXR, SHP, and SREBP-1c. J. Clin. Invest. 113,
1408–1418.
Woudenberg, J., Rembacz, K.P., van den Heuvel, F.A., Woudenberg-Vrenken,
T.E., Buist-Homan, M., Geuken, M., Hoekstra, M., Deelman, L.E., Enrich, C.,
Henning, R.H., et al. (2010). Caveolin-1 is enriched in the peroxisoma l mem-
brane of rat hepatocytes. Hepatology 51, 1744–1753.
Yang, X., Downes, M., Yu, R.T., Bookout, A.L., He, W., Straume, M., Mangels-
dorf, D.J., and Evans, R.M. (2006). Nuclear receptor expression links the circa-
dian clock to metabolism. Cell 126, 801–810.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors 247

Supplemental Information
EXTENDED EXPERIMENTAL PROCEDURES
Microarray Gene Expression Analysis
RNA Processing and Array Hybridization
Total RNA was analyzed for quality and quantity using the Agilent Bioanlayser and nanodrop ND-1000 spectrophotometer, respec-
tively. Biotin-labeled aRNA was synthesized with the Illumina RNA Amplification Kit (Ambion) as per the manufacturer’s instructions.
Briefly, 500 ng of total RNA was reverse transcribed to synthesize first- and second-strand cDNA, purified, and then in vitro tran-
scribed to synthesize biotin-labeled aRNA. A total of 1.5ng biotin-labeled cRNA was hybridized to each Illumina mouse WG6 v2 array
(Illumina) at 55
!
C for 18 hr. The hybridized BeadChip was washed and labeled with streptavidin-Cy3 (GE Healthcare) and then
scanned with the Illumina BeadStation 500 System (Illumina). The scanned image was imported into BeadStudio v2 software
(Illumina) and raw probe expression values were extracted.
Palmitate Oxidation in
k
CAV1+/+ and
k
CAV1"/" MEFs
Palmitate oxidation was determined by measuring tritiated water production using standard methods. Briefly, after serum with-
drawal, 500 ml of equilibrated (warm and gassed) assay media [high glucose DMEM (Invitrogen Australia) containing 2% fatty
acid-free BSA (Sigma-Aldrich), 0.25 mM potassium palmitate, and 1mCi/ml (20 nM) [9,10 (n)-
3
H]palmitic acid (GE Healthcare)] was
placed on cells. After 2 hr the media was removed, acidified with hydrochloric acid, and unoxidized palmitate removed via chloro-
form/methanol extraction. The aqueous phase containing titrated water was assayed in OptiPhase Supermix (PerkinElmer Life Sci-
ences, Waltham, MA) cocktail using the Wallac Trilux microbeta 1496 scintillation detector (PerkinElmer Life Sciences) and normal-
ized by protein concentration. Serial dilutions of [9,10 (n)-
3
H] palmitic acid mixed with water were used as activity standards.
Mouse Adipose Tissue Explant Culture
Adipose tissue was obtained from the epididymal fat pads of CAV1+/+ and CAV1 "/" mice between the ages of 14-18 weeks sacri-
ficed by CO
2
asphyxiation. Tissue was chopped into small pieces, washed, and maintained in DMEM medium supplemented with
0.1% fatty acid-free bovine serum albumin (PAA Laboratories, Austria) at 37
!
C for 2 hr prior to total RNA extraction.
CAV1 Knockdown in AML12-Hepatocytes
Transient CAV1 knockdown in AML12 hepatocytes using specific siRNA was performed as in (Ferna
´
ndez et al., 2006). Wild-type and
CAV1- knockdown AML12 hepatocytes were treated with DMSO or 10 mM wy14643 in F12/DMEM 1:1 (Invitrogen), 10% HISS for
24 hr prior to RNA extraction. Stable CAV1 and PTRF/Cavin1 knockdown AML12 cell lines were developed as described in (Ferna
´
n-
dez-Rojo et al., 2012). CAV1-kd that showed more than 95% reduction in CAV1 expression and #80% decrease in the number of
caveolae and PTRF/Cavin1-kd with a #60% reduction in caveolae density (data not shown) and still approximately expressed
#20% of the CAV1 protein levels observed in the WT AML12 cells.
Plasma Biochemical Analysis of BOH
Blood was extracted by cardiac puncture and collected in BD Biosciences microcontainer PST LH (#365987). Plasma BOH levels
were analyzed at the Clinical Pathology Laboratory, School of Veterinary Science (University Of Queensland).
Electron Microscopy
Liver samples were processed as in (Ferna
´
ndez et al., 2006) and (Ferna
´
ndez-Rojo et al., 2012).
Bioenergetic Metabolism in CAV1-Deficient AML12 Hepatocytes
Method as described in (Ferna
´
ndez-Rojo et al., 2012).
Quantitative Real-Time PCR and Primers
RNA was extracted using RNeasy (QIAGEN), and 4–5 mg was reverse transcribed. Quantitative RT-PCR was performed in triplicate
on 6-9 independent RNA preparations. cDNA levels were analyzed in PCR reactions with SYBR Green Technologies (Applied
Biosystems), and the relative level of expression was normalized using 18S ribosomal RNA. Primer sequences can be provided
on request.
Statistical Analysis
Statistical significance was assessed by using the Student’s t Test or one-way ANOVA in combination with Bonferroni’s multiple
comparison test unless otherwise indicated. For experiments on glycolysis and mitochondrial functions in CAV1-kd and PTRF/
Cavin1-kd AML12 hepatocytes significant differences among groups were defined at the conventional level p < 0.05 (two-tailed)
by general linear model (GLM), One-Way ANOVA test (SPSS 10 program for windows). ANOVA was used to evaluate the association
of groups in respect to diet (high glucose versus low glucose/Oleate) differences. T-paired tests were also used to study differences
within groups of high glucose and low glucose/Oleate medium, respectively. Significance is indicated as follows by (asterisks or
another symbol) *, p < 0.05; **, p < 0.01; ***, p < 0.001.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors S1

Figure S1. CAV1 Is a Major Regulator Essential for the Maintenance of Lipid Metabolism in Mice, Related to Figure 1
CAV1 deficiency in mice causes significant downregulation of genes involved in lipid metabolism function in MEFs, adipose tissue and liver. (See orange boxes).
Most importantly, the loss of CAV1 has a dramatic effect on lipid metabolism in liver despite hepatocytes being one of the mammalian cell types with lower
expression of CAV1. Red asterisk (*) highlights that IPA suggested that CAV1 deficiency has a significant effect on the expression of genes involved in pathways
related to methylation metabolism in MEFs, adipose tissue and liver. Bars graphs show the negati ve logarithm of the p = value for the most statistically significant
affected cellu lar functions in CAV1"/" tissues or MEFs relative to the clustering of the significantly affecte d genes per tissue.
S2 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

Figure S2. Lipodystrophy Is Not Responsible for the Impaired FAO and Ketogenesis Seen in Fasted CAV1"/" Mice, Related to Figure 1
(A) Plasma ketone body concentration in CAV1+/+ (white bar) and in vehicle (black bar), oleic acid (OA, dark gray bar) and arachidonic acid (AA, light gray bar)-
injected CAV1"/" mice.
(B and C) Gene expression of fatty acid oxidation-related PPARa-target genes (b) and of ketogenesis-related PPARa-target genes (c) in 12 weeks HFD-fed
CAV1+/+ (white bars) and CAV1"/" mice (black bars), n = 6.
(D) Plasma ketone body concentration in 12 week old HFD-fed CAV1+/+ and CAV1"/" mice.
(E) Heat-p lot representing a decrease (shades of blue) and an increase (shades of green) of the expression of genes in adipose tissue from 11 hr-fasted CAV1"/"
mice in comparison to adipose tissue from 11 hr-fasted CAV1+/+ mice (n = 3).
(F) Representative western-blots against PGC1a, PPARa and CAV1 in adipose tissue samples from individual mice at different times of fasting.
(G) PPARa, Cpt1b and MCAD gene expression analysis by qRT-PCR in adipose tissue explants from CAV1"/" (red bars) and CAV1+/+ (blue bars) mice. The data
represent the mean ± SEM. Statistical significance was assessed by using an Student’s t Test. Asterisks indicate *p < 0.05; **p < 0.01; ***p < 0.001.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors S3

Figure S3. The Effects of CAV1 in Hepatic Lipid Metabolism during Fasting Are Independent of Mouse Genetic Background and CAV1-
Independent Lack of Caveolae in PTRF/Cav in1"/" Impairs Liver Expression of FAO-Related Genes and Ketogenesis, Related to Figure 1
(A) Western blot in CAV1+/+ and CAV1"/" in liver homogenates from mice in a BalbC genetic background (
Balb/C
CAV1+/+ and
Balb/C
CAV1"/") shows that loss of
caveolin-1 reduced Caveolin-2, and PPARa protein levels.
(B)
Balb/C
CAV1"/" mice showed defective hepatic MCAD and PDK4 expression when compared to
Balb/C
CAV1+/+, similar to
K
CAV1"/" mice.
(C) As in
K
CAV1 mice, lack of CAV1 impairs concentrations of ketone bodies (b-Hydroxy butirate) in serum from 24 hr-fasted
Balb/C
CAV1 mice.
(D) Gene expression analysis of PPARa-target genes in 24 hr-fasted
JAX
CAV1+/+ and
JAX
CAV1"/" mice.
(E) As in
K
CAV1 mice, lack of CAV1 impairs concentrations of ketone bodies (b-Hydroxy butyrate) in serum from 24 hr-fasted
JAX
CAV1 mice.
(F) Liver CAV1 protein levels in fed (n = 4) and 24 hr-fasted (n = 5) PTRF/Cavin1+/+ and PTRF/Cavin1"/" mice. The data represent the mean ± SEM.
(G) Western blot against PPARa.
(H) Expression of liver fatty acid oxidation-related genes.
(I) Expression of liver ketogenesis-related genes.
(J) Plasma ketone bodies concentration. (B–E). The data represent the mean ± SEM (n = 5).
The data represent the mean ± SEM. Statistical significance was assessed by using a Student’s t Test. Asterisks indicate *p < 0.05; **p < 0.01; ***p < 0.001.
S4 Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors

Figure S4. Defective SHP Gene Regulation in CAV1"/" Livers, Related to Figure 4
(A) Estrogen receptor alpha expression from genome-wide expression analysis carried out in livers from fasted CAV1+/+ and CAV1"/" mice (n = 7).
(B) FXRa expression in fed and 24 hr-fasted CAV1+/+ and CAV1"/" mice (n = 8).
(C) Bile acid volumes in the gallbladders of 12 and 24 hr-fasted CAV1+/+ and CAV1"/" mice (n = 4-6).
(D) Serum bile acid concentration in 24 hr-fasted CAV1+/+ and CAV1"/" mice (n = 6).
The data represent the mean ± SEM. Statistical significance was assessed by using a Student’s t test. Asterisks indicate *p < 0.05; **p < 0.01; ***p < 0.001.
Cell Reports 4, 238–247, July 25, 2013 ª2013 The Authors S5