Article

High-Fat-Diet-Induced Obesity and Heart Dysfunction Are Regulated by the TOR Pathway in Drosophila

NASCR Center, Sanford/Burnham Medical Research Institute, La Jolla, CA 92037, USA.
Cell metabolism (Impact Factor: 17.57). 11/2010; 12(5):533-44. DOI: 10.1016/j.cmet.2010.09.014
Source: PubMed

ABSTRACT

High-fat-diet (HFD)-induced obesity is a major contributor to diabetes and cardiovascular disease, but the underlying genetic mechanisms are poorly understood. Here, we use Drosophila to test the hypothesis that HFD-induced obesity and associated cardiac complications have early evolutionary origins involving nutrient-sensing signal transduction pathways. We find that HFD-fed flies exhibit increased triglyceride (TG) fat and alterations in insulin/glucose homeostasis, similar to mammalian responses. A HFD also causes cardiac lipid accumulation, reduced cardiac contractility, conduction blocks, and severe structural pathologies, reminiscent of diabetic cardiomyopathies. Remarkably, these metabolic and cardiotoxic phenotypes elicited by HFD are blocked by inhibiting insulin-TOR signaling. Moreover, reducing insulin-TOR activity (by expressing TSC1-2, 4EBP or FOXO), or increasing lipase expression-only within the myocardium-suffices to efficiently alleviate cardiac fat accumulation and dysfunction induced by HFD. We conclude that deregulation of insulin-TOR signaling due to a HFD is responsible for mediating the detrimental effects on metabolism and heart function.

Full-text

Available from: Karen Ocorr
Cell Metabolism
Article
High-Fat-Diet-Induced Obesity and Heart Dysfunction
Are Regulated by the TOR Pathway in Drosophila
Ryan T. Birse,
1
Joan Choi,
1
Kathryn Reardon,
1
Jessica Rodriguez,
1
Suzanne Graham,
1
Soda Diop,
1
Karen Ocorr,
1
Rolf Bodmer,
1,
*
and Sean Oldham
1,
*
1
Development and Aging Program, NASCR Center, Sanford/Burnham Medical Research Institute, 10901 North Torrey Pines Road, La Jolla,
CA 92037, USA
*Correspondence: rolf@burnham.org (R.B.), soldham@burnham.org (S.O.)
DOI 10.1016/j.cmet.2010.09.014
SUMMARY
High-fat-diet (HFD)-induced obesity is a major
contributor to diabetes and cardiovascular disease,
but the underlying genetic mechanisms are poorly
understood. Here, we use Drosophila to test the
hypothesis that HFD-induced obesity and associ-
ated cardiac complications have early evolutionary
origins involving nutrient-sensing signal transduc-
tion pathways. We find that HFD-fed flies exhibit
increased triglyceride (TG) fat and alterations in
insulin/glucose homeostasis, similar to mammalian
responses. A HFD also causes cardiac lipid accu-
mulation, reduced cardiac contractility, conduction
blocks, and severe structural pathologies, reminis-
cent of diabetic cardiomyopathies. Remarkably,
these metabolic and cardiotoxic phenotypes eli-
cited by HFD are blocked by inhibiting insulin-TOR
signaling. Moreover, reducing insulin-TOR activity
(by expressing TSC1-2, 4EBP or FOXO), or
increasing lipase expression—only within the
myocardium—suffices to efficiently alleviate cardiac
fat accumulation and dysfunction induced by HFD.
We conclude that deregulation of insulin-TOR
signaling due to a HFD is responsible for mediating
the detr imental effects on metabolism and heart
function.
INTRODUCTION
Obesity has grown to epidemic proportions globally, with
more than 1.5 billion adults overweight and 400 million of
them considered obese. Increasing evidence indicates that
excessive dietary accumulation of lipids (i.e., ‘obesity’’) is
a high risk factor in causing deleterious effects on metabolism
and heart function and has been strongly linked to the
progression of heart disease and type 2 diabetes (Szendroedi
and Roden 2009; van Herpen and Schrauwen-Hinderling
2008). Investigating the origin and deleterious effects of
high-fat-diet (HFD)-induced obesity and its genetic mediators
is important in understanding the mechanisms that contribute
to obesity-associated secondary diseases, led by heart
disease.
However, the mechanisms that underlie HFD pathophysiology
have yet to be elucidated and include theories that the cause of
obesity is a recent evolutionary genetic adaptation to alternating
periods of famine and excess (Neel et al., 1998). Alternatively, the
potential for HFD-induced obesity may have arisen early in
evolution via nutrient-sensing pathways that coordinate metab-
olism. Here, we used the Drosophila model to investigate the
genetic mechanisms that may underlie HFD-induced obesity
and cardiac dysfunctions. Drosophila is well suited to study
this hypothesis. For example, the way the human heart is spec-
ified during embryogenesis relies on mechanisms and gene
networks that were first elucidated in Drosophila (Bodmer
1995; Harvey 1996), which has subsequently been validated in
many ways in mammals (Cripps and Olson 2002; Zaffran et al.,
2002; Ocorr et al., 2007a; Neely et al., 2010). Not only is the
embryology of cardiac specification equivalent between verte-
brates and invertebrates, but nearly the entire complement of
control genes and interactions is conserved and even partially
interchangeable between vertebrates and flies (Bodmer and
Venkatesh 1998; Cripps and Olson 2002). Since there is such
a deep evolutionary conservation in the formation of the heart,
it is likely that many aspects of heart function are also conserved,
including the mechanisms that control or influence how the heart
responds to HFD-induced obesity. In addition, Drosophila has
already been established as an excellent model for studying
the genetic control of metabolism and nutrient-sensing path-
ways (Baker and Thummel 2007; Leopold and Perrimon 2007;
Kim and Rulifson 2004; Oldham and Hafen 2003). The insulin
and TOR pathways are highly conserved regulators in the control
of metabolism. Although the molecular basis of regulating and
coordinating metabolic homeostasis is far from being under-
stood, manipulating insulin or TOR signaling in species ranging
from yeast to humans dramatically influences metabolic
responses, such as lipid and glucose homeostasis (Arking
et al., 2005; Saltiel and Kahn 2001; Tatar et al., 2003; Vellai
et al., 2003; Wang et al., 2005).
The genetic simplicity of the Drosophila model combined with
recently established heart function assays (Ocorr et al., 2007b,
2007c; Taghli-Lamallem et al., 2008; Wessells et al., 2004; Fink
et al., 2009) make it possible to probe the genetic mechanisms
of how a HFD affects heart function. Our findings indicate that
flies on a HFD become obese, which is dependent on insulin-
TOR signaling as well as regulators of lipid metabolisms (triacyl-
glycerol lipase and fatty acid synthase [FAS]). We show that both
systemic and (adipose) tissue-specific manipulations can protect
the fly from HFD-induced systemic lipid accumulation and other
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 533
Page 1
deleterious effects of high dietary fat, such as heart dysfunction.
Moreover, cardiac-specific inhibition of TOR downstream of
TSC1-2, overexpression of FOXO, or lipase protects the heart
dysfunction in otherwise obese flies due to exposure to HFD
conditions. Importantly, our evidence suggests that the meta-
bolic and cardiac dysfunction caused by a HFD is evolutionarily
and functionally conserved and mediated by a nutrient-sensitive
circuit, which includes insulin-TOR signaling and the control of
lipid metabolism. Thus, targeted inhibition of the TOR pathway
or lipid biogenesis may provide new therapeutic interventions
in obesity and its adverse effects on heart function.
RESULTS
Flies Exhibit Similar Phenotypes as Mammalian
Metabolic Syndrome
To investigate the conserved genetic mechanisms involved in
lipid accumulation in Drosophila, we first determined the effects
of a HFD on the fly’s metabolism. We found a dose-dependent
increase in whole-body TG content per body mass when flies
were fed diets containing 3%–30% saturated fats from coconut
oil for 5 days (Figure 1A, see Figure S1A available online). This
indicates that the fly accumulates lipids in a dose-dependent
wt-NF-2days
wt-HFD-2days
wt-NF-5days
wt-HFD-5days
wt-NF-10days
wt-HFD-10days
0.0
0.5
1.0
1.5
2.0
**
*
***
Glucose normalised to
wt NF flies
wt-NF
wt-HFD-2days
wt-NF-5days
wt-HFD-5days
wt-NF-10days
wt-HFD-10days
0
1
2
3
4
***
***
Relative change in Dilp2
wt-NF
wt-lauric-14% -5d
wt-myristic-5% 5d
wt-30%-5d
0.0
0.5
1.0
1.5
2.0
*
*
***
Triglycerides normalised
to wt NF flies
wt-NF
wt-3%-5d
wt-7%-5d
wt-15%-5d
wt-30%-5d
0.0
0.5
1.0
1.5
2.0
*
*
**
***
Triglycerides normalised
to wt NF flies
0%
20%
40%
60%
80%
100%
12345 12345
7cm
2-6cm
1cm
wt-NF
wt-HFD
Time (seconds)
AB
CD
E
Figure 1. HFD-Induced Obesity Leads to
Metabolic Syndrome in Flies
(A) TG content (expressed as relative change from
NF flies) of 10- to 15-day-old females on HFD for
5 days. The w
1118
fly strain was used as wild-
type (WT) in all experiments unless indicated
otherwise. WT flies were used under four different
concentrations of saturated fats. At least three
independent experiments were done for each
time point for all TG experiments. The concentra-
tions used were 3% (n = 46), 7% (n = 49), 15% (n
= 31), 30% (n = 81) for 5 days. WT type flies (n =
107) showed a dose-dependent increase in TGs
(*p < 0.05, **p < 0.01, ***p < 0.0001).
(B) TG content (expressed as relative change from
NF flies) of 10- to 15-day-old females on HFD for
5 days. The w1118 fly strain was used for the
wild-type (WT) flies. WT flies were used two dif-
ferent types of saturated fats. Myristic and lauric
acid are both the major components of coconut
oil. The amount of each fatty acid corresponds to
its respective amount found in the 30% HFD. At
least three independent experiments were done
for each time point for all TG experiments. The
concentrations used were lauric acid 14% (n =
36), while myristic acid was 5% (n = 35) and finally
30% of the orig inal coconut oil mixture for 5 days.
WT type flies showed an increase in TGs (*p < 0.05,
**p < 0.01, ***p < 0.0001).
(C) Glucose content of female WT flies fed a 30%
HFD (normali zed to WT NF-fed flies of each appro-
priate age). Trehalose present in the hemolymph
was converted to glucose (see the Experimental
Procedures). Since 30% HFD gave us the stron-
gest most consistent results, we used 30% for all
of the remaining experiments in this study. At least
three independent experiments were done for
each time point. After 2 days, glucose decreases,
then rapidly increases after 5 and 10 days
(*p < 0.05, **p < 0.01, ***p < 0.0001). A minimum
of 35 flies were used for each variable at each
time point.
(D) Relative Dilp2 transcript levels in WT flies fed
a 30% HFD for 2, 5, and 10 days. Dilp2 levels
rapidly increase after 2 days, then begin to
decrease the longer the fly remains on a HFD
(***p < 0.0001). All qPCR were done in triplicate.
(E) Graphical representation of the effects of HFD-
induced obesity on geotaxis. Flies were filmed for
5 s, then the movie was analyzed and individual
flies were counte d at each height, 1 cm being the lowest portion of the vial with 7 cm being the highest part of the vial. A minimum of 150 flies were used for
each variable. All flies eventually moved to the top of the vial, but we only counted the position of the flies to the allotted 5 s time span. Flies on a 30% HFD showed
a significant decrease in geotaxic activity, with approximately 80% remaining in the bottom of the vial for the allotted time span.
Error bars represent SEM.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
534 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 2
fashion. Next, we tested if the major individual fatty acids (FAs) in
the coconut oil could also cause an increase in TGs. We found
that either lauric or myristic acid (the main components of
coconut oil) caused a significant increase in whole body TGs
(Figure 1B, Figure S1B).
In mammals, increased levels of TGs have been found to be
a major risk factor for metabolic syndrome, insulin resistance,
and the onset of type 2 diabetes (Ouwens et al., 2005; Van Gaal
et al., 2006). In mammals after initial exposure to HFD, insulin
release is increased, which corresponds to a decrease in glucose
levels. However, upon chronic exposure to HFD, glucose levels
begin to increase as the periphery becomes more insulin resis-
tant. To determine if changes in insulin and glucose homeostasis
by HFD-induced obesity are functionally conserved, we tested
changes in glucose and insulin homeostasis in flies fed a HFD.
Indeed, after 2 days on a 30% HFD, flies showed a decrease
in total glucose accompanied by an increase in Drosophila
insulin-like peptide 2 (Dilp2) transcript levels (Figures 1C and
1D). After 5 and 10 days on HFD, glucose levels rose while
Dilp2 levels continued to drop (Figures 1C and 1D). Another
marker for monitoring insulin-glucose homeostasis is phosphor-
ylated Akt (pAkt), which decreases under HFD conditions,
thereby lowering glucose uptake in mammals (Manning, 2004).
In flies on a HFD, pAkt is progressively reduced (Figure S2A),
compared to flies on a normal food (NF) diet. We also find that
under a HFD the TOR pathway is activated, as measured by
increased 4EBP and S6K phosphorylation (Figure S2B).
Together, these data suggest that the Drosophila model exhibits
central aspects of metabolic syndrome and obesity upon HFD.
Excess dietary fat consumption has been linked with
increased nonadipose fat deposition, which is thought to be a
contributing factor in instigating several secondary diseases
such as type II diabetes, nonalcoholic fatty liver disease, colon
cancer, and cardiovascular disease (Schaffer 2003; Unger
2003; van Herpen and Schrauwen-Hinderling 2008). Therefore
we investigated the accumulation of lipids in both adult adipose
tissue (fat body, FB) and the midgut, which are the fly’s major
organs for lipid storage and utilization. Using oil red O to measure
the changes in fat accumulation, we find an increase in lipid
droplet staining in HFD-treated flies compared to NF fed controls
both in FB and gut (Figure S3). These findings indicate that when
fed a HFD flies accumulate excess fat in both adipose and non-
adipose tissue, as seen in mammals.
A recent study found that upon fat accumulation activity levels
tend to decrease (Wareham, 2007). Therefore, we determined the
activity levels of flies under HFD conditions. Normally flies prefer
to move against gravity (negatively geotaxic) (Gargano et al.,
2005). When testing for changes in geotaxic activity, we found
that HFD-fed flies exhibit a substantial reduction in geotaxic
behavior (Movies S1 and S2), with most flies remaining at the
bottom of the vial during the experimental time (Figure 1E).
They will eventually climb to the top, albeit more slowly and
less vigorously. These data suggest that consumption of a HFD
has adverse, lethargic-like effects on the flies’ activity levels.
Increased Dietary Fat Causes Severe Functional
and Structural Changes in the Fly Heart
When rodents are fed a HFD, they show signs of increased
accumulation of cardiac TGs, accompanied by hypertrophy
and decreases in fractional shortening (FS), a relative measure
of cardiac contractility (Fang et al., 2008; Sowers, 2003). It
may be the increased adipose and circulating lipid levels that
cause a disturbance of cardiac performance. For example,
increased expression of lipid transporters in the heart leads to
elevated fat in the myocardial cells accompanied by cardiac
dysfunction (Chiu et al., 2005), although the underlying mecha-
nisms have yet to be elucidated. To determine if ‘fat’ flies
also show increased cardiac TG levels and exhibit deteriorating
heart function, we examined them under HFD conditions. After
5 days on a HFD, we observed elevated cardiac TG levels (Fig-
ure 2A) and progressively altered contraction patterns (Figures
2B–2G, Movies S3–S5). We then investigated other cardiac
dysfunctions and observed a reduced FS due to a smaller dia-
stolic
diameter (Figures 4C and 4D), reminiscent of a restrictive
cardiac phenotype (see Cammarato et al., 2008). In addition, we
found noncontractile portions of the heart, dysfunctional inflow
valves (ostia), and increased incidences of partial conduction
blocks (anterior and posterior heart beats at a different rate)
(Figures 2D–2G, Movies S6–S8). We also found severe defects
and disorganization in the myofibrillar structure of the heart of
flies fed a HFD (Figures 2H1–2H2
00
). Taken together, these
results dramatically illustrate that a HFD in Drosophila adversely
affects heart function, and that in both mammals and flies a HFD
compromises cardiac activity and structural integrity.
Systemic Inhibition of the TOR Pathway Prevents
Excess Fat Accumulation
It is unclear how obesity increases the risk of heart disease. Since
the TOR pathway has been linked to nutrient sensing and the
modulation of aging heart function in Drosophila (Luong et al.,
2006; Wessells et al., 2009), we tested whether TOR signaling
is involved in the regulation of lipid levels related to cardiac
dysfunction. To determine if reducing TOR function can
alter the effects of HFD-induced obesity, we fed hypomorphic
TOR
7/P
mutant flies (Luong et al., 2006) a HFD and observed no
increase in TG levels, compared to flies on NF (Figure 3A). The
TOR mutants show lower TG levels on both NF and a HFD,
compared to wild-type on NF (Figure S1C). These results suggest
that reducing TOR function is either accelerating fat catabolism or
decreasing lipid anabolism or storage. To test possible down-
stream mechanisms that may contribute to this phenotype, we
tested the change in transcript levels of genes involved in lipid
metabolism: Brummer (ATGL) lipase (Bmm) (Gronke et al.,
2005) and fatty acid synthase (FAS) (Valet et al., 2002). We found
that Bmm transcript levels are significantly increased, whereas
FAS levels are decreased, in TOR
7/P
mutants under both NF and
HFD conditions (Figures 3B and 3C), suggesting that TOR activity
regulates the balance between lipid anabolism and catabolism,
possibly due to increased lipase and decreased FAS activity.
We tested this idea further by using oil red O, a lipid droplet
marker, to investigate the changes in fat accumulation in both
the FB and gut. Indeed, HFD-fed TOR mutants showed no
increase in the quantity of lipid droplets in both the FB and gut,
and fat levels were generally lower (Figure S3).
To determine if TOR-dependent prevention of HFD-induced
obesity extends to other HFD-mediated phenotypes, we tested
the geotactic activity of the TOR mutants under NF and HFD
conditions. We find that in TOR mutants these activity levels
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 535
Page 3
0% 20% 40% 60% 80% 100%
wt-30%-HFD
wt-15%-HFD
wt-7%-HFD
wt-3%-HFD
wt-NF
Partial Conduction
Block
Non-contractile
Dysfunctional Ostia
No noticeable
defects
Heart Triglycerides
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
0
1
2
3
*
Triglycerides normalised
to wt NF flies
Heart period
wt-NF
wt-3%-HFD
wt-7%-HFD
wt-15%-HFD
wt-30%-HFD
0.0
0.2
0.4
0.6
0.8
***
***
***
***
Seconds
wt-NF
wt-30%
-5 days
wt-15%
-5days
ABC
H
D
F
E
G
Figure 2. HFD Treatment Causes Cardiac Dysfunction
(A) TG content from female hearts on NF and HFD for 5 days (normalized to WT NF hearts). TOR
7/P
mutant flies were examined under the same food conditions.
At least three independent experiments were done for each time point for all TG experiments (with approximately 90 hearts were used for each condition and
genotype). WT type flies showed an increase in TGs (*p < 0.05), while the TOR mutants had significantly lower levels of TGs that did not increase on a HFD.
(B) Bar graph representation of changes in heart period for a population of approximately 24 flies’ hearts from each food type. A dose dependent effect is shown
for the increase in dietary fat content (***p < 0.0001). A minimum of 22 flies were used for each dietary condition.
(C) M-mode traces prepared from high-speed movies of semi-intact preparations on NF, and 15% and 30% HFDs for 5 days. Note the changes in the M-mode
traces as the concentration of fat increases. Red bars indicate systolic and diastolic diameters.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
536 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 4
are already impaired under NF, and this phenotype is exacer-
bated when flies are placed on a HFD (Figure 3D), which
suggests that the protection by the ubiquitous inhibition of
TOR activity is not a global response but may be selective to
tissue type and lipid metabolism. These data are consistent
with mice that are deficient in TOR complex 1 (TORC1) in skeletal
muscles, which exhibit a 40% decrease in voluntary wheel
activity (Bentzinger et al., 2008 ).
Systemic Inhibition of the TOR Pathway Protects
the Heart against HFD-Induced Cardiac Dysfunction
Next, we tested whether global reduction of TOR function pro-
tects the heart’s performance under HFD conditions. First, we
tested whether the elevated TGs due to a HFD are reduced in
the heart itself in TOR
7/P
mutants. Indeed, reduced systemic
TOR function causes a significant decrease in cardiac TGs,
which remains low even under HFD conditions similar to the
(D) A sketch of a dissected semi-intact heart preparation. Arrows indicate to the area the corresponding M-modes shown in (E)–(F
0
).
(E–E
0
) Representation of partial conduction blocks of hearts using M-modes from different portions of the same heart. (E) M-mode from the anterior portion of
the heart displaying a regular heart beat. (E
0
) M-mode from the posterior portion of the same heart displaying a faster and erratic heart beating pattern, com-
pared to I.
(F–F
0
) Representation of a portion of the heart that is noncontractile using M-modes of different heart regions. (F) M-mode taken from the anterior portion of the
heart displays regular beating pattern. (F
0
) M-mode taken from the posterior portion of the same heart showing poor or no contractions.
(G) Side bar graph of heart phenotypes seen under HFD conditions with increasing amount of fat. The heart phenotypes used in this graph are partial conduction
blocks, noncontractile myocardial cells, dysfunctional ostia, and no noticeable defects. The instances of all three phenotypes increase in a fat-dose-dependent
fashion. A minimum of 26 heart movies were analyzed for each dietary condition.
(H) Fluorescent micrographs of hearts on NF and HFD for 5 days at 103 and 253 optical magnification. Adult hearts are stained with Alexa Fluor 594 phalloidin. (1)
Adult heart of a WT flies on NF. (1
0
) Magnified anterior NF-fed WT heart region. Note the regular arrangements of myofibrillar organization. (2) Adult heart of a WT
flies on HFD for 5 days. Note the degeneration of the regular myofibrillar structure and the decrease in heart tube diameter. (2
0
) Magnified anterior WT heart region
on HFD. Note the disorganization of the circular myocardial myofibrils. (3) Adult heart of a TOR
7/P
mutant on a HFD for 5 days. Note that there is little to no degra-
dation of heart structure, and the diameter is not constricted. (3
0
) Magnified anterior heart region of a TOR
7/P
mutant. There is little change in myocardial structure
under HFD conditions compared to NF-fed WT (1
0
).
Error bars represent SEM.
CA
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
12345 12345
7cm
2-6cm
1cm
TOR7/P-NF TOR7/P-HFD
Percent of flies at designated heights
Time (seconds)
D
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
0
1
2
3
4
***
***
**
Relative change in Lipase
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
0
1
2
3
***
***
Expression of FAS
B
wt-NF
wt-HFD
-HFD
7/P
TOR
arm>TSC -HFD
0.0
0.5
1.0
1.5
2.0
***
Triglycerides normalised
to NF flies
Figure 3. Reducing TOR Function Prevents
HFD Obesity
(A) Comparison of changes in TG levels (ex-
pressed as relative change from NF flies) between
WT, TOR
7/P
, and systemic TSC1-2 overexpression
(arm > TSC1-2 flies). At least three independent
experiments were performed for each time point
for all TG experiments. A significant increase in
TGs were seen in WT flies (***p < 0.0001) fed a
30% HFD for 5 days, while TOR mutants (n = 46)
started with lower levels of TG under NF, and these
levels did not increase even on 30% HFD (n = 48).
Arm > TSC flies (n = 30) had simila r levels as the
WT control flies, but these levels did not increase
when placed on a HFD.
(B) Relative Bmm (ATGL Lipase) mRNA transcript
levels. There is a slight increase in Lipase levels
in WT under 30% HFD conditions (p < 0.05).
In contrast, TOR
7/P
mutants have a 3-fold in-
crease in Bmm levels under NF conditions and
remain high under HFD conditions. All qPCRs
were done in triplicate.
(C) Expression of FAS transcript levels in WT and
TOR
7/P
mutants under both NF and 30% HFD con-
ditions for 5 days. TOR
7/P
flies show significantly
lower FAS mRNA levels than the corresponding
WT flies on the same diets (**p < 0.01). All qPCR
were done in triplicate.
(D) Graphical representation of the effects of HFD-
induced obesity on geotaxis. Flies were filmed for
5 s, then the movi e was analyzed and individual
flies were counted at each height, 1 cm being the
lowest portion of the vial, with 7 cm being the high-
est part of the vial. A minimum of 150 flies were
used for each variable. All flies eventually moved
to the top of the vial, but we only counted the posi-
tion of the flies to the allotted 5 s time span. Unlike
WT, TOR
7/P
are impaired in their geotaxic ability
that was exac erbated when fed a HFD.
Error bars represent SEM.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 537
Page 5
results from whole-body TGs (Figure 2 A). We then show that the
TOR mutants exhibit considerable protection against deteriora-
tion of heart function when placed on a HFD. Remarkably, flies
with reduced TOR function exhibit unchanged heartbeat and
contractility parameters under NF or HFD, comparable to wild-
type under NF conditions (Figures 4A–4E, Movie S9). Similarly,
the incidences of conduction blocks, dysfunctional ostia, and
noncontractile myocardial regions were as low as in the TOR
mutants under NF or HFD conditions compared to wild-type flies
on NF (Figure 4F). We also find that the myocardial structure
remains intact and similar to WT on NF (Figures 2H3–2H3
0
).
These findings suggest that a decrease in TOR activity initiates
lipid breakdown and decreased lipid synthesis, which may
help prevent the adverse effects of a HFD on the heart.
Metabolic homeostasis is a balance between storage and
expenditure that is regulated by genetic and environmental
interactions, both polygenic and multiorgan in nature. To eluci-
date how reduced TOR signaling alters the HFD response at
the tissue level, we used the UAS/Gal4 system (Brand and Per-
rimon, 1993) to inhibit TOR pathway components both in the
TOR7/P
-30%-5days
wt-NF
wt-30%
-5 days
arm>TSC
-30%-5day
A
BC
D
E
Heart period
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
arm>TSC -NF
arm>TSC-HFD
0.0
0.2
0.4
0.6
0.8
1.0
***
Seconds
Fractional Shortening
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
arm>TSC -NF
arm>TSC-HFD
0.0
0.1
0.2
0.3
0.4
0.5
*
%
Diastolic Diameter
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
arm>TSC -NF
arm>TSC-HFD
0
20
40
60
80
*
Microns
Systolic Diameter
wt-NF
wt-HFD
-NF
7/P
TOR
-HFD
7/P
TOR
arm>TSC -NF
arm>TSC-HFD
0
10
20
30
40
50
Microns
0% 20% 40% 60% 80% 100%
arm>TSC-HFD
arm>TSC-NF
TOR7/P-HFD
TOR7/P-NF
wt-HFD
wt-NF
Partial Conduction Block
Non-contractile
Dysfunctional Ostea
No noticeable defects
F
Figure 4. Reducing TOR Function Prevents HFD-Induced Obesity Cardiac Dysfunction
(A) M-mode traces of dissected WT flies on NF and 30% HFD, TOR
7/P
mutants and arm > TSC1-2 on NF and 30% HFD. No significant change was seen in TOR
mutant nor arm < TSC heart M-modes on HFD when compared to WT on NF.
(B) Bar graph of cumulative heart periods for WT, TOR mutants (n = 25, 36 respectively) and arm > TSC1-2 flies (n = 20 [NF], 21 [HFD]) under NF and HFD condi-
tions. No change in heart period from WT under NF was seen in TOR mutants and arm > TSC1-2 flies under HFD conditions.
(C) Bar graph of changes in FS. A decrease in FS in WT flies was seen after 5 days on a 30% HFD (p < 0.001). While no change was seen in FS in neither the TOR
mutants nor the arm > TSC1-2 flies under a HFD.
(D) Bar graph of combined diastolic diameter data of WT and TOR mutant and arm > TSC1-2 hearts under NF and 30% HFD conditions. No change was seen in
diastolic diameter in TOR mutants and arm > TSC1-2 flies under a HFD.
(E) Bar graph of systolic diameter of fly heart as in (F). No decrease in systolic diameter was seen after 5 days on a 30% HFD for any of the strains tested.
(F) Graphical representation of heart phenotypes as in Figure 2 G. No significant change to the heart phenotype of TOR
7/P
mutants or with systemic TSC1-2 over-
expression under HFD could be detected, when compared to WT flies. A minimum of 20 individual fly heart movies were analyzed.
Error bars represent SEM.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
538 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 6
whole organism and in selective tissues. First, we overexpressed
the TOR inhibitor TSC1-2 ubiquitously using the arm-Gal4 driver,
which resulted in a similar rescue of the HFD-induced increase in
TGs and aberrant heart phenotypes (Figure 4), as with TOR loss-
of-function mutants. Moreover, global inhibition of TOR transla-
tional effectors (eIf4E and S6K) by 4EBP (Miron et al., 2001) or
dominant-negative S6K overexpression (S6K
DN
)(Hennig and
Neufeld, 2002), respectively, also elicits a protective cardiac
phenotype against HFD-induced abnormalities (data not shown),
similar to TSC1-2 overexpression (Figure 4). These results indi-
cate that systemic inhibition of TOR signaling protects the animal
against the adverse cardiac effects of a HFD and that multiple
TOR-dependent effectors may contribute to TG accumulation
and cardiac dysfunction.
Adipose-Specific Manipulation of Insulin-TOR Signaling
Affects Obesity Phenotypes
In insects, the FB is an important organ for lipid storage and
hormonal regulation and is functionally comparable to the verte-
brate liver and adipose tissue. Therefore, we tested whether inhi-
bition of TOR signaling in the fly’s FB can prevent the HFD-
induced obesity and heart phenotypes. Expression of TSC1-2,
S6K
DN
, and 4EBP specifically in the FB showed a phenotype
similar to the ubiquitous inhibition of TOR exposed to a HFD.
Specifically, FB-restricted expression abolished the normally
observed increase in TGs, and there are no significant changes
in cardiac function of all these genotypes when subjected to
HFD conditions (Figure 5, Figure S4). Thus, genetic manipula-
tions inhibiting TOR pathway activation specifically in the FB
resulted in protection from HFD-induced fat accumulation and
cardiac dysfunction. This result also suggests that manipulation
of TOR in the FB regulates the organism’s systemic lipid metab-
olism, which in turn influences proper heart function.
Since we find that TOR mutants exhibit an increase in Bmm
and a decrease in FAS transcript levels (Figures 3B and 3C),
we tested if expression of Bmm or FAS-RNAi in the FB protects
against HFD-induced increases in body fat and heart dysfunc-
tion. Indeed, we find that expression of Bmm in the FB prevents
systemic TG accumulation and protects against abnormalities in
heart function under HFD, similar to FB expression of TSC1-2
(Figure 5, Figure S4). When we tested the effects of FAS knock-
down in the FB, we found a partial decrease in TG levels and
a moderate attenuation of HFD-associated heart defects (Fig-
ure 5, Figure S4). These findings suggest that prevention of fat
accumulation by increasing lipolysis or reducing lipogenesis
can mimic the metabolic and cardiac protection from HFD
observed by reduced TOR signaling.
It has been shown that loss of insulin signaling in the murine
adipose leads to lower circulating insulin levels and prevents
obesity (FIRKO mouse; Bluher et al., 2002). Therefore we tested
if altering insulin signaling in the FB had physiological conse-
quences similar to TOR inhibition. To achieve this, we expressed
the negative effector of insulin signaling, FOXO, in the FB. This
also limited the accumulation of lipids, as we observed for TOR
inhibition and Bmm lipase expression, and protected the heart
from the adverse effects of excess lipid accumulation (Figure 5,
Figure S4). Thus, reduced insulin-TOR signaling in the FB
prevents systemic dyslipidemia and protects the heart (and
presumably other organs) from lipid overload.
Autonomous Protection of the Heart from HFD
by Cardiac Inhibition of Insulin-TOR Signaling
The above findings raise the question of whether manipulating
insulin-TOR signaling in the heart itself is sufficient to protect
the heart against the effects of a HFD. To test this, we expressed
TSC1-2, 4EBP, and S6K
DN
specifically in the myocardial cells
of the heart and found that the systemic TG levels show the
same increase when fed a HFD, as did wild-type flies (Figure 6A,
Figure S5A). Remarkably, we also observed that myocardial
expression of both 4EBP and S6K
DN
was sufficient to protect
against HFD-induced cardiac abnormalities, despite elevated
systemic lipid levels (Figure 6, Figure S5B). However, unlike all
other TOR inhibitory manipulations, overexpression of TSC1-2
in the heart did not confer a robust protection from a HFD-
inflicted insult to the heart (Figure 6, Figure S5B). To further
confirm that inhibition of TOR itself was sufficient to prevent
HFD-induced cardiac dysfunction, we expressed a dominant-
negative form of TOR, TOR
TED
(Hennig and Neufeld, 2002), in
the heart and found similar cardiac protection, as with 4EBP
and S6K
DN
, against the adverse effects of a HFD (Figure 6),
comparable to the systemic or FB-specific manipulations
(Figure 5).
To determine if elevated lipase activity is a possible mecha-
nism downstream of reduced TOR activity in the protection of
the heart, we directly expressed the Bmm lipase in the myocar-
dial cells of flies exposed to a HFD. We found that increased
cardiac lipase expression protected the hearts against HFD-
associated dysfunction, yet the systemic TG levels were
increased under a HFD, as in wild-type (Figure 6A, Figure S5B).
We also tested heart-specific FAS knockdown and observed
similar protective phenotypes to those observed with Bmm
overexpression or inhibition of TOR signaling downstream of
TSC1-2 (Figure 6, Figure S5B). These findings strongly imply
that reduced cardiac TOR signaling autonomously protects the
heart’s susceptibility to a HFD by altering lipid metabolism.
Since adipose FOXO expression protects against HFD-
induced cardiac malfunction (Figures 5B and 5C) and cardiac
FOXO expression protects against cardiac aging (Wessells
et al., 2004), we also tested whether FOXO activity in the heart
under HFD conditions is also protective. Indeed, we find that
the expression of FOXO in myocardial cells autonomously
protects the heart from the adverse effects of a HFD, but as
expected not against overall body fat accumulation (Figure 6A).
We then tested if altering downstream insulin-TOR signaling in
the heart tissue would affect lipid accumulation with the heart.
We found no significant increase in heart-specific TG accumula-
tion in HFD hearts when overexpressing either FOXO or Bmm
(Figure S5C). We also found that under NF conditions Bmm over-
expression had significantly lower levels of heart TGs under
NF conditions then WT controls (Figure S5C). These findings
correlate well with our previous data showing a decrease in TG
accumulation when inhibiting insulin-TOR or increasing lipase
activity. Taken together, these results suggest that moderate
reduction in insulin-TOR signaling prevents HFD-induced
obesity and cardiac dysfunction (Figure 7). It will be interesting
to see whether FOXO acts upstream of TOR effectors (i.e.,
4EBP) in protecting the heart from lipotoxicity, as it does in
reducing the decline of cardiac performance with age (Wessells
et al., 2009).
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 539
Page 7
DISCUSSION
HFD-induced obesity is associated with an increased risk for
diseases, including cancer, diabetes, and heart disease. The
polygenic and/or multiorgan nature of HFD-induced obesity
makes it difficult to determine the relative contribution to each
of these diseases. In order to examine these complicated inter-
actions in a simple system, we established a HFD-induced
obesity model in Drosophila to elucidate the underlying mecha-
nisms. We used the genetic versatility of the Drosophila model
along with sophisticated cardiac function assays to investigate
the effects of insulin-TOR-mediated metabolic regulation and
the crosstalk between organs exposed to excess dietary fat.
We find that HFD-fed flies become obese, develop metabolic
syndrome, and exhibit severe symptoms of cardiac lipotoxicity.
The deleterious HFD-induced effects are alleviated by genetic
manipulations of metabolic regulators or by directly altering lipid
metabolism either systemically, in adipose tissue, or specifically
(autonomously) in the heart.
We provide evidence that Drosophila fed a HFD exhibit central
features of mammalian metabolic syndrome, including elevated
lipid levels and changes in insulin and glucose homeostasis.
BA
Relative Triglyceride levels
wt-NF
wt-HFD
lsp>TSC -HFD
-HFD
DN
lsp>S6K
lsp>4EBP -HFD
lsp>FOXO -HFD
lsp>Bmm -HFD
lsp>FAS-RNAi -HFD
0.0
0.5
1.0
1.5
2.0
***
**
**
Triglycerides normalised
to NF flies
Heart Dysfunction
wt-NF
wt-HFD
lsp>TSC-HFD
-HFD
DN
lsp>S6K
lsp>4EBP-HFD
lsp>FOXO-HFD
lsp>Bmm-HFD
lsp>FAS-RNAi-HFD
0
1
2
3
4
5
**
***
Relative change in heart
dysfunction
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
lsp>FAS-RNAi-HFD
lsp>FAS-RNAi-NF
lsp>Bmm-HFD
lsp>Bmm-NF
lsp>FOXO-HFD
lsp>FOXO-NF
lsp>4EBP-HFD
lsp>4EBP-NF
lsp>S6K-DN-HFD
lsp>S6K-DN-NF
lsp>TSC-HFD
lsp>TSC-NF
wt-HFD
wt-NF
Partial Conduction Block
Non-contractile
Dysfunctional Ostia
No noticeable defects
C
Figure 5. Fat Body-Specific Inhibition of TOR Prevents Obesity and Heart Phenotypes
(A) Changes in TG content of WT and FB-specific (lsp-Gal4) expression of TSC1-2 (n = 36), S6K
DN
(n = 36), 4EBP (n = 34), FOXO (n = 48), Bmm (n = 36), and FAS-
RNAi (n = 34). WT flies show a significant increase in TGs, while all of the other flies show no significant increase in TG levels after 5 days on 30% HFD.
(B) Bar graph of relative change in heart dysfunction for FB-specific (lsp-Gal4) expression of TSC1-2 (n = 24), S6K
DN
(n = 36), 4EBP (n = 24), UAS-FOXO (n = 36),
UAS-Bmm (n = 36), and UAS-FAS-RNAi (n = 24). Only FAS-RNAi had a moderate increase in the severity of all three cardiac dysfunction phenotypes (see C) when
compared to the same genotype under HFD.
(C) Side bar graph of individual aberrant heart phenotypes for WT and FB-specific expression of TSC1-2, S6K
DN
, 4EBP, FOXO; FAS-RNAi; and Bmm. The only
strain to exhibit a moderate increase in heart abnormalities under a HFD was FAS-RNAi. A minimum of 24 individual fly heart movies were analyzed for each
genotype and diet variation.
Error bars represent SEM.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
540 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 8
Furthermore, the effects of the HFD-induced obesity on the heart
are profound, including diminished contractility, conduction
blocks, and structural defects. The HFD-induced elevation in TG
levels is an important marker for this collection of phenotypes, as
high TG levels are associated with disruptions of lipid and glucose
homeostasis, mitochondrial function, and other processes
(Schaffer 2003; Unger 2003; Ouwens et al., 2005; Van Gaal et al.,
2006; van Herpen and Schrauwen-Hinderling 2008), all of which
may contribute to high lipid accumulation and heart phenotypes.
Because the insulin-TOR pathway is a key integrator of metab-
olism, we initiated a comprehensive investigation of both the
systemic and tissue-specific effects of altering insulin-TOR
signaling under HFD conditions. We found that reduction of
pathway activity blocks HFD-induced increased lipid levels in
Drosophila. Recent studies have begun to connect insulin-TOR
signaling to the regulation of lipid metabolism in flies and
mammals (Luong et al., 2006; Li et al., 2010; Lee et al., 2010;
this study). For example, the increased lipid synthesis caused
by insulin treatment is blocked by reduction of TOR function,
and the activation of the TOR pathway leads to fat accumulation
(Luong et al., 2006; Porstmann et al., 2008; Li et al., 2010).
In addition, TOR function in lipogenesis (and heart function)
AB
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
GMH5>FAS-RNAi-HFD
GMH5>FAS-RNAi-NF
GMH5>Bmm-HFD
GMH5>Bmm-NF
GMH5>4EBP-HFD
GMH5>4EBP-NF
GMH5>S6K-DN-HFD
GMH5>S6K-DN-NF
GMH5>TOR-TED-HFD
GMH5>TOR-TED-NF
GMH5>TSC-HFD
GMH5>TSC-NF
wt-HFD
wt-NF
Partial Conduction Block
Non-contractile
Dysfunctional Ostia
No noticeable defects
C
Heart Dysfunction
wt-NF
wt-HFD
GMH5>TSC-HFD
-HFD
TED
GMH5>TOR
-HFD
DN
GMH5>S6K
GMH5>4EBP-HFD
GMH5>FOXO-HFD
GMH5>Bmm-HFD
GMH5>FAS-RNAi-HFD
0
1
2
3
4
5
***
**
Relative change in heart
dysfunction
Relative Triglyceride levels
wt-NF
wt-HFD
GMH5>TSC-HFD
-HFD
TED
GMH5>TOR
-HFD
DN
GMH5>S6K
GMH5>4EBP-HFD
GMH5>FOXO-HFD
GMH5>Bmm-HFD
GMh5>FAS-RNAi-HFD
0.0
0.5
1.0
1.5
2.0
2.5
***
Triglycerides normalised
to NF flies
Figure 6. Myocardial-Specific Inhibition of TOR Autonomously Blocks HFD Heart Effects Despite Obesity
(A) Relative changes in TG content of WT and myocardial-specific expression (GMH5-Gal4) of TSC1-2 (n = 36), TOR
Ted
(n = 36), S6K
DN
(n = 48), 4EBP (n = 36),
FOXO (n = 35), Bmm (n = 48), and FAS-RNAi (n = 36). All flies tested show a significant increase in TGs levels after 5 days on 30% HFD.
(B) Bar graph of relative change in heart dysfunction for myocardial-specific (GMH5) expression of TSC1-2 (n = 23), TOR
TED
(n = 24), S6K
DN
(n = 35), 4EBP (n = 42),
FOXO (n = 22), Bmm (n = 33) and FAS-RNAi (n = 25). Only TSC1-2 had an increase in the incidences of all three phenotypes when compared to the same genotype
under HFD.
(C) Side bar graph of individual aberrant heart phenotypes for myocardial-specific (GMH5) expression of TSC1-2, TOR
TED
, S6K
DN
, 4EBP, FOXO, and Bmm. Only
TSC1-2 had an increase in the incidences of all three phenotypes when compared to the same genotype under HFD. A minimum of 22 individua l fly heart movies
were analyzed for each genotype and diet variation.
Error bars represent SEM.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 541
Page 9
may be mediated by activation of sterol regulatory element
binding protein (SREBP) and its lipogenic target genes, including
FAS (H.Y. Lim, R.T.B., S.O., and R.B., unpublished data). There-
fore, modulation of lipid metabolism is a likely mechanism that
mediates the effect of TOR signaling on the HFD phenotype.
To determine the contributions of lipid metabolism to the HFD-
induced obesity phenotypes, we examined a key regulator of
lipid utilization: the Bmm (ATGL) gene is required for the break-
down of lipid droplets, and encodes a triacylglycerol lipase that
is conserved from nematodes to mammals (Gronke et al.,
2005, 2007; this study). Drosophila Bmm mutants show signifi-
cant increases in TGs, while ectopic expression results in the
reverse effect of lowering TG levels. Previous studies in mice
have shown that systemic mutants of ATGL caused increases
in lipid accumulation in nonadipose such as the heart (Haem-
merle et al., 2006; Hirano et al., 2008). Here, we find that reduc-
tion of TOR function leads to increased Bmm RNA levels and that
Bmm overexpression prevents the HFD-induced elevation of TG
levels. In the heart, Bmm overexpression protects against
cardiac dysfunction and fat accumulation inflicted by a HFD.
These protective effects by Bmm may stem from modified lipid
utilization by the mitochondria. Additionally, lipid droplet lipolysis
may also liberate fatty acid ligands for nuclear receptors, which
may contribute to changes in mitochondrial biogenesis (Palanker
et al., 2009; Finck et al., 2003). Altered insulin-TOR signaling is
likely to also increase activity of translation factors involved in
mitochondrial function (Zid et al., 2009). Lastly, increases in
autophagy have been implicated in the regulation of lipid metab-
olism via the breakdown of lipid droplets (Kovsan et al., 2009).
Thus, reduction of insulin-TOR signaling can coordinately lead
to changes in lipid metabolism, which in turn affects organismal
physiology under excess dietary fat conditions.
A critical step in the regulation of lipid synthesis involves
SREBP, which in flies responds to decreased levels of the major
membrane lipid phosphatidylethanolamine (PE) by increasing
FAS expression (Rawson, 2003). In examining the role of lipid
synthesis in relation to the HFD-induced obesity phenotypes,
we observed that the transcript levels of FAS decrease in TOR
mutants, and that FAS knockdown reduces the deleterious
effects of HFD, possibly because of reduced lipid synthesis.
Consistent with this idea, loss of SREBP and FAS function in
mammalian models leads to low TG levels (Valet et al., 2002;
Bentzinger et al., 2008), and loss of Drosophila easily shocked
(eas), which encodes an ethanolamine kinase critical for PE
synthesis, leads to increased levels of the active form of SREBP,
increased FAS expression, and thus elevated TG levels (H.Y. Lim
and R.B., unpublished data). In addition, this effect is also
observed in mammalian hepatocytes where TOR function is
required for SREBP activation (Li et al., 2010), and it has been
proposed that TOR serves as an important fork in the road of
diabetic insulin resistance separating gluconeogenesis from lipo-
genesis (Li et al., 2010; Laplante and Sabatini, 2010). Collectively,
these data support the idea that TOR is required to mediate HFD-
induced obesity and ensuing (cardiac) organ defects via multiple
mechanisms. Interestingly, reducing or blocking fat accumula-
tion mimics the protective effects of lowered insulin-TOR
signaling under a HFD. However, it remains to be determined
what it is about the accumulating fat that is detrimental to an
organism: is the fat itself or a side effect of its accumulation that
is causing metabolic and physiological dysregulation?
Our data indicate that heart dysfunction due to the HFD treat-
ment is the result of autonomous changes within the heart, as
evidenced by the increased cardiac TG levels; and cardiac-
only reduction of insulin-TOR signaling protects the heart from
dysfunction and fat accumulation. Importantly, our findings
show that inhibition of insulin-TOR signaling or ensuing fat accu-
mulation in the heart itself can significantly prevent cardiac
dysfunction despite the presence of elevated systemic TG
levels. In addition, HFD-induced obesity also influences heart
function via the adipose tissue, since blocking insulin-TOR func-
tion in the adipose can also prevent the HFD obesity heart
phenotypes. Thus, nonautonomous crosstalk factors, which
may include hormones or metabolites, also contribute to HFD-
induced heart dysfunction. The Drosophila model
will help to
understand the basis of such crosstalk.
The fact that flies become obese on a HFD has important impli-
cations on its own. Theories for the increased frequency and
appearance of obesity include the thrifty gene hypothesis, which
tries to understand why the incidence of obesity may be
increasing despite its deleterious effects. These ideas state that
recent evolutionary selection pressure on people experiencing
frequent famines concentrated genetic variants that increased
the ability to provide sufficient nutrients to an organism in times
of lean, but also predisposed to obesity in times of plenty. Alter-
natively, the potential for HFD-induced obesity may have arisen
early in evolution, perhaps independently of external selection
pressure, via deregulation of metabolic responses in multicellular
organisms. In this report, we provide evidence to suggest that the
capacity for HFD-induced obesity and its associated complica-
tions is likely evolutionarily ancient and inherent to core
nutrient-sensing pathways, a property that manifests itself
upon exposure to dietary extremes resulting in a homeostatic
imbalance that exceeds the normal dynamic range.
Figure 7. Model of Insulin-TOR Pathway Function in HFD-Induced
Obesity
Model for the effects of increased lipids on the insulin-TOR pathway in the FB
and the heart.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
542 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 10
The discovery of these HFD-induced obesity phenotypes in
the Drosophila genetic model will permit a detailed dissection
of obesity phenotypes, especially with regard to the cardiac lip-
otoxicity effects (and possibly mimicking aspects of diabetic
cardiomyopathy). In particular, we can now attempt to under-
stand the various contributions of insulin resistance, fat accumu-
lation, and fatty acid oxidation to the HFD-induced obesity
phenotypes, including timing requirements. In summary, the
advent of the Drosophila HFD-induced obesity model opens up
many new horizons to study deregulated processes and
diseases of chronic lipid excess.
EXPERIMENTAL PROCEDURES
Fly Stocks
We obtained w
1118
flies from Bloomington’s Stock Center and used these as
wild-type controls. We used multiple drivers for each tissue type. For driving
expression ubiquitously, we used arm-Gal4 (Bloomington); in the FB we
used the lsp-Gal4 (Cherbas et al., 2003) and the DCG-Gal4 (provided by Jon
Graff and J. Suh); for the heart we used GMH5 (Wessells et al., 2004) and
Hand-Gal4 (Han et al., 2006). We found similar trends for corresponding
drivers for each tissue. S6K
DN
(Mary Stewart), and UAS-d4EBP (Miron et al.,
2001), UAS-FAS-RNAi (Vienna RNAi stock center), and UAS-Bmm (Gronke
et al., 2005) were used. UAS-TOR
TED
and UAS-TSC1-2 were a donation (by
Tom Neufeld). The TOR mutants are from Luong et al. (2006). To test a negative
effector of the insulin signaling pathway, we used UAS-FOXO (Wessells et al.,
2004). All flies were maintained on NF source made from a combination of
yeast, corn starch, and molasses. HFD was made by mixing either 3%, 7%,
15%, or 30% coconut oil to the food in a weight-to-volume ratio with the NF.
Heart Assays
For a detailed analysis of the cardiac contrac tions it is necessary to surgically
expose the fly’s heart in order to make it accessible for high-resolution video
microscopy (Vogler and Ocorr 2009). In this analysis we used the previously
published heart assay and analysis (Ocorr et al., 2007a, 2007b, 2007c; Fink
et al., 2009). In brief, dissections to expose the fly’s heart within the abdomen
were kept in oxygenated saline, and high-speed digital movies were taken and
then analyzed for heart rate, arrhythmia, contractility, FS, etc. (Ocorr et al.,
2007a, 2007b, 2007c; Fink et al., 2009).
HFD Feeding Regime
Bottles of flies were emptied and dated; then 5 days after emptying the vial, all
flies were taken and placed in a new vial of NF and aged 5 more days. This pop-
ulation was then spilt into two populations, one on NF and one on the desig-
nated concentration of HFD, for either 2, 5, or 10 days. We used 30% HFD
for 5 days for the majority of the experiments, since it gave strong and repro-
ducible phenotypes.
Statistical Analysis
All statistical analysis was done using Student’s t tests, and the analysis was
performed using GraphPad Prism version 5.00 for Windows, (GraphPad Soft-
ware, San Diego California USA, http://www.graphpad.com/).
Additional methods are in the Supplemental Information available online.
SUPPLEMENTAL INFORMATION
Supplemental Information includes five figures, nine movies, Supplemental
Experimental Procedures, and Supplemental References and can be found
with this article at doi:10.1016/j.cmet.2010.09.014.
ACKNOWLEDGMENTS
We would like to thank J. Graff, J. Suh, Laurent Perrin, M. Stewart, Tom Neu-
feld, and N. Sonenberg for reagents and fly stocks. We would also like to thank
Daniel Kelly, Timothy Osborne, and Michael Karin for critical reading of the
manuscript. We thank Lisa Elme
´
n for technical assistance and fly stock
curating. R.T.B. was supported by fellowships from CIRM and from the San-
ford Child Health Center at the Sanford-Burnham Institute. K.O. was sup-
ported by a Scientist Development Grant of the American Heart Association
(AHA). S.O. was supported by AHA and NHLBI of NIH. R.B. was supported
by NHLBI of NIH, MDA, and the Ellison Foundation. R.T.B designed and per-
formed experiments, analyzed data, and wrote the paper; J.C., K.R., J.R.,
S.G., and S.D. performed experiments; K.O. developed analytical tools and
helped analyze data; and R.B. and S.O. supervised the project, analyzed
data, and wrote the paper.
Received: May 7, 2010
Revised: July 13, 2010
Accepted: August 10, 2010
Published: November 2, 2010
REFERENCES
Arking, D.E., Atzmon, G., Arking, A., Barzilai, N., and Dietz, H.C. (2005). Asso-
ciation between a functional variant of the KLOTHO gene and high-density
lipoprotein cholesterol, blood pressure, stroke, and longevity. Circ. Res. 96,
412–418.
Baker, K.D., and Thummel, C.S. (2007). Diabetic larvae and obese flies-
emerging studies of metabolism in Drosophila. Cell Metab. 6, 257–266.
Bentzinger, C.F., Romanino, K., Cloetta, D., Lin, S., Mascarenhas, J.B., Oliveri,
F., Xia, J., Casanova, E., Costa, C.F., Brink, M., et al. (2008). Skeletal muscle-
specific ablation of raptor, but not of rictor, causes metabolic changes and
results in muscle dystrophy. Cell Metab. 8, 411–424.
Bluher, M., Michael, M.D., Peroni, O.D., Ueki, K., Carter, N., Kahn, B.B., and
Kahn, C.R. (2002). Adipose tissue selective insulin receptor knockout protects
against obesity and obesity-related glucose intolerance. Dev. Cell 3, 25–38.
Bodmer, R. (1995). Heart development in Drosophila and its relationship to
vertebrate systems. Trends Cardiovasc. Med. 5, 21–28.
Bodmer, R., and Venkatesh, T.V. (1998). Heart development in Drosophila and
vertebrates: conservation of molecular mechanisms. Dev. Genet. 22, 181–186.
Brand, A.H., and Perrimon, N. (1993). Targeted gene expression as a means of
altering cell fates and generating dominant phenotypes. Development 118,
401–415.
Cammarato, A., Dambacher, C.M., Knowles, A.F., Kronert, W.A., Bodmer, R.,
Ocorr, K., and Bernstein, S.I. (2008). Myosin transducer mutations differentially
affect motor function, myofibril structure, and the performance of skeletal and
cardiac muscles. Mol. Biol. Cell 19, 553–562.
Cherbas, L., Hu, X., Zhimulev, I., Belyaeva, E., and Cherbas, P. (2003). EcR iso-
forms in Drosophila: testing tissue-specific requirements by targeted blockade
and rescue. Development 130, 271–284.
Chiu, H.C., Kovacs, A., Blanton, R.M., Han, X., Courtois, M., Weinheimer, C.J.,
Yamada, K.A., Brunet, S., Xu, H., Nerbonne, J.M., et al. (2005). Transgenic
expression of fatty acid transport protein 1 in the heart causes lipotoxic cardio-
myopathy. Circ. Res. 96, 225–233.
Cripps, R.M., and Olson, E.N. (2002). Control of cardiac development by an
evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28.
Fang, C.X., Dong, F., Thomas, D.P., Ma, H., He, L., and Ren, J. (2008). Hyper-
trophic cardiomyopathy in high-fat diet-induced obesity: role of suppression of
forkhead transcriptio n factor and atrophy gene transcription. Am. J. Physiol.
Heart Circ. Physiol.
295,
H1206–H1215.
Finck, B.N., Han, X., Courtois, M., Aimond, F., Nerbonne, J.M., Kovacs, A.,
Gross, R.W., and Kelly, D.P. (2003). A critical role for PPARalpha-mediated
lipotoxicity in the pathogenesis of diabetic cardiomyopathy: modulation by
dietary fat content. Proc. Natl. Acad. Sci. USA 100, 1226–1231.
Fink, M., Callol-Massot, C., Chu, A., Ruiz-Lozano, P., Izpisua Belmonte, J.C.,
Giles, W., Bodmer, R., and Ocorr, K. (2009). A new method for detection and
quantification of heartbeat parameters in Drosophila, zebrafish, and embry-
onic mouse hearts. Biotechniques 46, 101–113.
Gargano, J.W., Martin, I., Bhandari, P., and Grotewiel, M.S. (2005). Rapid iter-
ative negative geotaxis (RING): a new method for assessing age-related loco-
motor decline in Drosophila. Exp. Gerontol. 40, 386–395.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc. 543
Page 11
Gronke, S., Mildner, A., Fellert, S., Tennagels, N., Petry, S., Muller, G., Jackle,
H., and Kuhnlein, R.P. (2005). Brummer lipase is an evolutionary conserved fat
storage regulator in Drosophila. Cell Metab. 1, 323–330.
Gronke, S., Muller, G., Hirsch, J., Fellert, S., Andreou, A., Haase, T., Jackle, H.,
and Kuhnlein, R.P. (2007). Dual lipolytic control of body fat storage and mobi-
lization in Drosophila. PLoS Biol. 5, e137. 10.1371/journal.pbio.0050137.
Haemmerle, G., Lass, A., Zimmermann, R., Gorkiewicz, G., Meyer, C., Roz-
man, J., Heldmaier, G., Maier, R., Theussl, C., Eder, S., et al. (2006). Defective
lipolysis and altered energy metabolism in mice lacking adipose triglyceride
lipase. Science 312, 734–737.
Han, Z., Yi, P., Li, X., and Olson, E.N. (2006). Hand, an evolutionarily conserved
bHLH transcription factor required for Drosophila cardiogenesis and hemato-
poiesis. Development 133, 1175–1182.
Harvey, R.P. (1996). NK-2 homeobox genes and heart development. Dev. Biol.
178, 203–216.
Hennig, K.M., and Neufeld, T.P. (2002). Inhibition of cellular growth and prolif-
eration by dTOR overexpression in Drosophila. Genesis 34, 107–110.
Hirano, K., Ikeda, Y., Zaima, N., Sakata, Y., and Matsumiya, G. (2008). Triglyc-
eride deposit cardiomyovasculopathy. N. Engl. J. Med. 359, 2396–2398.
Kim, S.K., and Rulifson, E.J. (2004). Conserved mechanisms of glucose
sensing and regulation by Drosophila corpora cardiaca cells. Nature 431,
316–320.
Kovsan, J., Bashan, N., Greenberg, A., and Rudich, A. (2009). Potential role of
autophagy in modulation of lipid meta bolism. Am. J. Physiol. Endocrinol.
Metab. 298, E1–E7.
Laplante, M., and Sabatini, D.M. (2010). mTORC1 activates SREBP-1c and
uncouples lipogenesis from gluconeogenesis. Proc. Natl. Acad. Sci. USA
107, 3281–3282.
Lee, J.H., Budanov, A.V., Park, E.J., Birse, R., Kim, T.E., Perkins, G.A., Ocorr,
K., Ellisman, M.H., Bodmer, R., Bier, E., and Karin, M. (2010). Sestrin as a feed-
back inhibitor of TOR that prevents age-related pathologies. Science 327,
1223–1228.
Leopold, P., and Perrimon, N. (2007). Drosophila and the genetics of the
internal milieu. Nature 450, 186–188.
Li, S., Brown, M.S., and Goldstein, J.L. (2010). Bifurcation of insulin signaling
pathway in rat liver: mTORC1 required for stimulation of lipogenesis, but not
inhibition of gluconeogenesis. Proc. Natl. Acad. Sci. USA 107, 3441–3446.
Luong, N., Davies, C.R., Wessells, R.J., Graham, S.M., King, M.T., Veech, R.,
Bodmer, R., and Oldham, S.M. (2006). Activated FOXO-mediated insulin resis-
tance is blocked by reduction of TOR activity. Cell Metab. 4, 133–142.
Manning, B.D. (2004). Balancing Akt with S6K: implications for both metabolic
diseases and tumorigenesis. J. Cell Biol. 167, 399–403.
Miron, M., Verdu, J., Lachance, P.E., Birnbaum, M.J., Lasko, P.F., and Sonen-
berg, N. (2001). The translational inhibitor 4E-BP is an effector of PI(3)K/Akt
signalling and cell growth in Drosophila. Nat. Cell Biol. 3, 596–601.
Neel, J.V., Weder, A.B., and Julius, S. (1998). Type II diabetes, essential hyper-
tension, and obesity as ‘syndromes of impaired genetic homeostasis’’: the
‘thrifty genotype’ hypothesis enters the 21st century. Perspect. Biol. Med.
42, 44–74.
Neely, G.G., Kuba, K., Cammarat o, A., Isobe, K., Amann, S., Zhang, L., Murata,
M., Elme
´
n, L., Gupta, V., Arora, S., et al. (2010). A global in vivo Drosophila
RNAi screen identifies NOT3 as a conserved regulator of heart function. Cell
141, 142–153.
Ocorr, K., Akasaka, T., and Bodmer, R. (2007a). Age-related cardiac disease
model of Drosophila. Mech. Ageing Dev. 128, 112–116.
Ocorr, K., Reeves, N.L., Wessells, R.J., Fink, M., Chen, H.S., Akasaka, T.,
Yasuda, S., Metzger, J.M., Giles, W., Posakony, J.W., and Bodmer, R.
(2007b). KCNQ potassium channel mutations cause cardiac arrhythmias in
Drosophila that mimic the effects of aging. Proc. Natl. Acad. Sci. USA 104,
3943–3948.
Ocorr, K.A., Crawley, T., Gibson, G., and Bodmer, R. (2007c). Genetic variation
for cardiac dysfunction in Drosophila. PLoS ONE 2, e601. 10.1371/journal.
pone.0000601.
Oldham, S., and Hafen, E. (2003). Insulin/IGF and target of rapamycin
signaling: a TOR de force in growth control. Trends Cell Biol. 13, 79–85.
Ouwens, D.M., Boer, C., Fodor, M., de Galan, P., Heine, R.J., Maassen, J.A.,
and Diamant, M. (2005). Cardiac dysfunction induced by high-fat diet is asso-
ciated with altered myocardial insulin signalling in rats. Diabetologia 48, 1229–
1237.
Palanker, L., Tennessen, J.M., Lam, G., and Thummel, C.S. (2009). Drosophila
HNF4 regulates lipid mobilization and beta-oxidation. Cell Metab. 9, 228–239.
Porstmann, T., Santos, C.R., Griffiths, B., Cully, M., Wu, M., Leevers, S., Grif-
fiths, J.R., Chung, Y.L., and Schulze, A. (2008). SREBP activity is regulated by
mTORC1
and contributes to Akt-dependent cell growth. Cell Metab. 8, 224–
236.
Rawson, R.B. (2003). The SREBP pathway—insights from Insigs and insects.
Nat. Rev. Mol. Cell Biol. 4, 631–640.
Saltiel, A.R., and Kahn, C.R. (2001). Insulin signalling and the regulation of
glucose and lipid metabolism. Nature 414, 799–806.
Schaffer, J.E. (2003). Lipotoxicity: when tissues overeat. Curr. Opin. Lipidol.
14, 281–287.
Sowers, J.R. (2003). Therapeutic lifestyle changes in the management of
obesity. Endocr. Pract. 9 (Suppl 2 ), 97–98.
Szendroedi, J., and Roden, M. (2009). Ectopic lipids and organ function. Curr.
Opin. Lipidol. 20, 50–56.
Taghli-Lamallem, O., Akasaka, T., Hogg, G., Nudel, U., Yaffe, D., Chamberlain,
J.S., Ocorr, K., and Bodmer, R. (2008). Dystrophin deficiency in Drosophila
reduces lifespan and causes a dilated cardiomyopathy phenotype. Aging
Cell 7, 237–249.
Tatar, M., Bartke, A., and Antebi, A. (2003). The endocrine regulation of aging
by insulin-like signals. Science 299, 1346–1351.
Unger, R.H. (2003). Minireview: weapons of lean body mass destruction: the
role of ectopic lipids in the metabolic syndrome. Endocrinology 144, 5159–
5165.
Valet, P., Tavernier, G., Castan-Laurell, I., Saulnier-Blache, J.S., and Langin, D.
(2002). Understanding adipose tissue developme nt from transgenic animal
models. J. Lipid Res. 43, 835–860.
Van Gaal, L.F., Mertens, I.L., and De Block, C.E. (2006). Mechanisms linking
obesity with cardiovascular disease. Nature 444, 875–880.
van Herpen, N.A., and Schrauwen-Hinderling, V.B. (2008). Lipid accumulation
in non-adipose tissue and lipotoxicity. Physiol. Behav. 94, 231–241.
Vellai, T., Takacs-Vellai, K., Zhang, Y., Kovacs, A.L., Orosz, L., and Muller, F.
(2003). Genetics: influence of TOR kinase on lifespan in C. elegans. Nature
426, 620.
Vogler, G., and Ocorr, K. (2009). Visualizing the beating heart in Drosophila. J.
Vis. Exp. 10.3791/1425.
Wang, P., Zhang, Y., Li, H., Chieu, H.K., Munn, A.L., and Yang, H. (2005). AAA
ATPases regulate memb rane association of yeast oxyst erol binding proteins
and sterol metabolism. EMBO J. 24, 2989–2999.
Wareham, N.J. (2007). Epidemiological studies of physical activity and dia-
betes risk, and implications for diabetes prevention. Appl. Physiol. Nutr.
Metab. 32, 778–782.
Wessells, R.J., Fitzgerald, E., Cypser, J.R., Tatar, M., and Bodmer, R. (2004).
Insulin regulation of heart function in aging fruit flies. Nat. Genet. 36, 1275–
1281.
Wessells, R., Fitzgerald, E., Piazza, N., Ocorr, K., Morley, S., Davies, C., Lim,
H.Y., Elmen, L., Hayes, M., Oldham, S., and Bodmer, R. (2009). d4eBP acts
downstream of both dTOR and dFoxo to modulate cardiac functional aging
in Drosophila. Aging Cell 8, 542–552.
Zaffran, S., Xu, X., Lo, P.C., Lee, H.H., and Frasch, M. (20 02). Cardiogenesis in
the Drosophila model: control mechanisms during early induction and diversi-
fication of cardiac progenitors. Cold Spring Harb. Symp. Quant. Biol. 67, 1–12.
Zid, B.M., Rogers, A.N., Katewa, S.D., Vargas, M.A., Kolipinski, M.C., Lu, T.A.,
Benzer, S., and Kapahi, P. (2009). 4E-BP extends lifespan upon dietary restric-
tion by enhancing mitochondrial activity in Drosophila. Cell 139, 149–160.
Cell Metabolism
A High-Fat Diet Obesity Model in Drosophila
544 Cell Metabolism 12, 533–544, November 3, 2010 ª2010 Elsevier Inc.
Page 12
  • Source
    • "Upd2, Ilp6 (discussed above)] or an increase in glucose-stimulated secretion from the IPCs secondary to the reduction in glucose disposal by the fat body remains unknown. Insulin-resistant Drosophila have also been generated by rearing flies on high-sugar diet (HSD) (Morris et al., 2012; Musselman et al., 2011; Skorupa et al., 2008) or high-fat diet (HFD) (Birse et al., 2010). Similar to InR 05545 /InR + insulin-resistant flies, HSD causes insulin resistance with ILP compensation. "
    [Show abstract] [Hide abstract] ABSTRACT: Mechanisms of glucose homeostasis are remarkably well conserved between the fruit fly Drosophila melanogaster and mammals. From the initial characterization of insulin signaling in the fly came the identification of downstream metabolic pathways for nutrient storage and utilization. Defects in these pathways lead to phenotypes that are analogous to diabetic states in mammals. These discoveries have stimulated interest in leveraging the fly to better understand the genetics of type 2 diabetes mellitus in humans. Type 2 diabetes results from insulin insufficiency in the context of ongoing insulin resistance. Although genetic susceptibility is thought to govern the propensity of individuals to develop type 2 diabetes mellitus under appropriate environmental conditions, many of the human genes associated with the disease ingenome-wide association studies have not been functionally studied. Recent advances in the phenotyping of metabolic defects have positioned Drosophila as an excellent model for the functional characterization of large numbers of genes associated with type 2 diabetes mellitus. Here, we examine results from studies modeling metabolic disease in the fruit fly and compare findings to proposed mechanisms for diabetic phenotypes in mammals. We provide a systematic framework for assessing the contribution of gene candidates to insulin-secretion or insulinresistance pathways relevant to diabetes pathogenesis.
    Preview · Article · Apr 2016 · Disease Models and Mechanisms
  • Source
    • "More recently, the key metabolic genes modulating cardiac lipotoxicity were described, showing the protective role of Drosophila Peroxisome proliferator-activated receptor γ coactivator-1 (PGC1), called spargel (srl), downstream of the TOR pathway [147]. The deleterious high calorie diets–induced cardiac dysfunction is alleviated by down-regulating brummer, the fly homolog of adipocyte triglyceride lipase (ATGL), and spargel, suggesting that cardiac-specific inhibition of TOR blocks the heart effects of a high caloric diet in flies [143,146,147]. Likewise, the induction of SREBP by TOR activation, in parallel to ATGL inhibition, caused FAS induction and lipid storage. "
    Full-text · Article · Feb 2016
  • Source
    • "In a Drosophila model of diet-induced heart disease, dietary lipid content was increased by the addition of coconut oil, a lipid-rich (undefined) ingredient [9]. Minimal organismal effects were seen at concentrations below 30% coconut oil [9]. Considering that I was able to increase dietary lipid content four-fold without any noticeable adverse effects on lifespan or developmental timing, one can estimate that, in order to bring about the kinds of changes observed in " high-fat " / " high-lipid " diets based on 30% coconut oil, dietary lipid content must be increased by considerably more than four-fold, and/or the added source of lipids must contain higher proportions of specific storage-promoting molecules (e.g., TAGs). "
    [Show abstract] [Hide abstract] ABSTRACT: Gene-diet interactions play a crucial but poorly understood role in susceptibility to obesity. Accordingly, the development of genetically tractable model systems to study the influence of diets in obesity-prone genetic backgrounds is a focus of current research. Here I present a modified synthetic Drosophila diet optimized for timely larval development, a stage dedicated to energy storage. Specifically increasing the levels of individual macronutrients-carbohydrate, lipid, or protein-resulted in markedly different organismal effects. A high-carbohydrate diet adversely affected the timing of development, size, early lifespan and body fat. Strikingly, quadrupling the amount of dietary lipids had none of these effects. Diets rich in protein appeared to be the most beneficial, as larvae developed faster, with no change in size, into long-lived adults. I believe this synthetic diet will significantly facilitate the study of gene-diet interactions in organismal energy balance.
    Full-text · Article · Jan 2016 · PLoS ONE
Show more