Mitofusin 2 in POMC Neurons Connects
ER Stress with Leptin Resistance
and Energy Imbalance
Marc Schneeberger,1,2,3Marcelo O. Dietrich,4,5David Sebastia ´n,3,6,7Mo ´nica Imberno ´n,8,9Carlos Castan ˜o,1,3
Ainhoa Garcia,1,3Yaiza Esteban,1,3Alba Gonzalez-Franquesa,1Ignacio Castrillo ´n Rodrı ´guez,3,6,7Analı ´a Bortolozzi,10,11
Pablo M. Garcia-Roves,1,3Ramon Gomis,1,2,3Ruben Nogueiras,8,9Tamas L. Horvath,4Antonio Zorzano,3,6,7
and Marc Claret1,3,*
1Diabetes and Obesity Research Laboratory, Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), 08036 Barcelona, Spain
2Department of Endocrinology and Nutrition, Hospital Clı ´nic. School of Medicine, University of Barcelona, 08036 Barcelona, Spain
3Centro de Investigacio ´n Biome ´dica en Red de Diabetes y Enfermedades Metabo ´licas Asociadas (CIBERDEM), 08036 Barcelona, Spain
New Haven, CT 06520, USA
5Department of Biochemistry, Universidade Federal do Rio Grande do Sul, Porto Alegre RS 90035, Brazil
6Institute for Research in Biomedicine (IRB Barcelona), 08028 Barcelona, Spain
7Departament de Bioquı ´mica i Biologia Molecular, Facultat de Biologia, Universitat de Barcelona, 08028 Barcelona, Spain
8Department of Physiology, School of Medicine, University of Santiago de Compostela–Instituto de Investigacio ´n Sanitaria, 15782 Santiago
de Compostela, Spain
9Centro de Investigacio ´n Biome ´dica en Red Fisiopatologı ´a de la Obesidad y Nutricio ´n (CIBERobn), 15782 Santiago de Compostela, Spain
10Department of Neurochemistry and Neuropharmacology, IIBB–CSIC–IDIBAPS, 08036 Barcelona, Spain
11Centro de Investigacio ´n Biome ´dica en Red de Salud Mental (CIBERSAM), 08036, Barcelona, Spain
Mitofusin 2 (MFN2) plays critical roles in both
mitochondrial fusion and the establishment of mito-
chondria-endoplasmic reticulum (ER) interactions.
Hypothalamic ER stress has emerged as a causative
factor for the development of leptin resistance, but
the underlying mechanisms are largely unknown.
Here, we show that mitochondria-ER contacts in
anorexigenic pro-opiomelanocortin (POMC) neurons
in the hypothalamus are decreased in diet-induced
obesity. POMC-specific ablation of Mfn2 resulted
in loss of mitochondria-ER contacts, defective
POMC processing, ER stress-induced leptin resis-
tance, hyperphagia, reduced energy expenditure,
and obesity. Pharmacological relieve of hypotha-
Our data establish MFN2 in POMC neurons as an
essential regulator of systemic energy balance by
fine-tuning the mitochondrial-ER axis homeostasis
and function. This previously unrecognized role for
MFN2 argues for a crucial involvement in mediating
ER stress-induced leptin resistance.
The mediobasal hypothalamus is a critical area of the central
nervoussystemimplicatedintheregulation ofhomeostatic func-
tions. Specific populations of neurons in the arcuate nucleus
(ARC) play fundamental roles in the regulation of energy balance
(Williams and Elmquist, 2012). In particular, neurons coexpress-
ing orexigenic neuropeptides agouti-related protein (AgRP) and
neuropeptide Y (NPY) along with neurons coexpressing anorex-
igenic pro-opiomelanocortin(POMC) precursor and cocaine and
amphetamine-related transcript (CART) have been extensively
involved in the regulation of appetite, body weight, and meta-
bolism (Die ´guez et al., 2011; Myers and Olson, 2012; Sisley
and Sandoval, 2011; Williams and Elmquist, 2012).
A hallmark of obesity is leptin resistance, in which high-circu-
lating leptin levels are unable to promote its central anorexigenic
effects (Bjørbaek, 2009). POMC neurons are primary targets of
leptin in the hypothalamus and evidence suggests that alter-
ations in this subset of neurons may mediate the development
of hypothalamic leptin resistance (Bjørbaek, 2009; Diano et al.,
2011; Gamber et al., 2012). Recent data demonstrate that
enhanced hypothalamic endoplasmic reticulum (ER) stress and
activation of the unfolded protein response (UPR) play a primary
pathogenic role in leptin resistance development. For example,
both diet-induced and genetic models of obesity have been
associated withincreased ERstressinthe hypothalamus (Ozcan
et al., 2009; Won et al., 2009; Zhang et al., 2008). Pharmacolog-
ical or genetic ER stress induction causes hypothalamic leptin
resistance and, consequently, increased appetite and obesity
(Hosoi et al., 2008; Ozcan et al., 2009; Won et al., 2009; Zhang
et al., 2008). Chemical chaperones relieve hypothalamic ER
stress and act as leptin sensitizers, reducing food intake and
body weight (Ozcan et al., 2009; Won et al., 2009; Zhang et al.,
2008). Despite the strong link between ER stress and leptin
172 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
resistancein thehypothalamus, thecellularandmolecular deter-
minants involved remain elusive.
The ER and mitochondria form contacts that support essential
structural and functional interorganelle communication, a pro-
cess that is partially modulated by mitochondrial fusion and
fission (Rowland and Voeltz, 2012). Recent evidences indicate
that Mitofusin 2 (MFN2), a ubiquitously expressed mitochondrial
transmembrane dynamin-like GTPase protein, plays funda-
mental roles in both mitochondrial fusion (Chen et al., 2003)
and the establishment of mitochondria-ER interactions (de Brito
tions have been associated with the development of metabolic
alterations and recent data suggest a key role for MFN2 in the
pathophysiology of obesity and type 2 diabetes in both humans
and rodents (Bach et al., 2005; Bach et al., 2003; Herna ´ndez-
Alvarez et al., 2009; Herna ´ndez-Alvarez et al., 2010; Sebastia ´n
et al., 2012).
These evidences encouraged us to hypothesize that MFN2 in
POMC neurons may be a molecular link between ER stress
and leptin resistance in the hypothalamus. Here, we provide
evidence that mitochondrial dynamics and mitochondria-ER
interactions in POMC neurons are altered in diet-induced
obesity (DIO). Our data also indicate that Mfn2 deficiency in
this population of neurons causes loss of mitochondria-ER con-
tacts, ER stress, leptin resistance, and reduced a-melanocyte
stimulating hormone (a-MSH) production, leading to severe
obesity. These results unravel an unexpected role for MFN2
in the regulation of POMC neuronal function and hypotha-
lamic ER stress to appropriately regulate whole-body energy
Mitochondrial Network and Mitochondria-ER Contacts
in POMC Neurons Are Altered in DIO
The cellular adaptations to fluctuations in nutrient availability
involve the regulation of mitochondrial function, a process that
is intimately associated with changes in mitochondrial dynamics
and mitochondria-ER interactions (Baltzer et al., 2010; Mandl
et al., 2009; Youle and van der Bliek, 2012). We assessed poten-
tial alterationsin POMC neuron
complexity caused by energy excess in DIO mice using confocal
microscopy. Aspect ratio (AR) and form factor (FF), parameters
associated with mitochondria length and branching, respec-
tively, were decreased in POMC neurons from DIO mice when
compared to lean controls (Figures 1A, 1B, S1A, and S1B avail-
able online). The positive correlation between mitochondria AR
and FF in POMC neurons from lean mice was lost in DIO mice
(Figure 1C). To further examine the effects of nutrient excess in
POMC neurons, we conducted electron microscopy (EM) anal-
ysis. As previously described, mitochondrial density in POMC
neurons from DIO mice was increased (Diano et al., 2011; data
not shown). Interestingly, DIO mice showed a significant reduc-
tion in the number of mitochondria-ER contacts in POMC
neurons when compared to lean controls (Figures 1D and 1E).
Collectively, these results indicate that both mitochondrial
network complexity and their association to ER in POMC neu-
rons are altered in DIO mice.
Reduced Mfn2 Expression in the Hypothalamus
Precedes the Onset of Obesity and Its Overexpression
Ameliorates the DIO Phenotype
A candidate protein to mediate the alterations in mitochondria-
ER contacts and mitochondrial network observed in DIO mice
is MFN2. To initially assess this, we conducted a time course
study and analyzed the expression of genes implicated in mito-
chondrial dynamics in the hypothalamus of lean and DIO mice at
different time points. Although 4 days of high-fat diet (HFD) did
not cause changes in body weight (Figure 1F), hypothalamic
Mfn2 expression was already downregulated (Figure 1G).
Reduced Mfn2 expression in the hypothalamus was observed
throughout the study (Figure 1G). Expression of mitofusin 1
(Mfn1), optic atrophy 1 (Opa1), dynamin-related protein 1
(Drp1), and mitochondrial fission 1 (Fis1) remained unchanged
sion of MFN2 in the ARC of DIO mice was able to improve their
metabolic alterations. Localization studies showed correct
delivery of the adenovirus (Figure 2A). MFN2 overexpression
(Figure 2B) in the ARC of DIO mice resulted in reduced body
ure 2F) and increased interscapular surface temperature
adjacent to brown adipose tissue (BAT) (Figure 2G). Consistent
uncoupling protein 1 (Ucp1) was upregulated (Figure 2H), while
the trend in hypothalamic neuropeptide expression was
congruent with increased anorexigenic output (Figure 2I).
Remarkably, MFN2 overexpression in the ARC attenuated the
hypothalamic DIO-induced expression of ER stress markers
tein (Chop), spliced form of X-box-binding protein 1 (Xbp1s),
activating transcription factor 4 (Atf4) and 6 (Atf6) (Figure 2J).
Together, these results suggest that MFN2 in the hypothala-
mus may underlie the metabolic alterations of DIO mice, and
hence it can be a relevant protein implicated in the central regu-
lation of energy balance.
BAT thermogenic marker
Generation of Mice Lacking MFN2 in POMC Neurons
MFN2 immunoreactivity was widely observed in the ARC
including POMC neurons (Figure S2A). To explore the role of
MFN2 in POMC neurons upon the regulation of whole-body en-
ergy homeostasis, we generated POMC neuron-specific Mfn2
knockout mice (POMCMfn2KO). We confirmed tissue-specific
recombination of the Mfn2-floxed allele in the hypothalamus
(Figure S2B). We also detected reduction in Mfn2 gene expres-
sion but not in other mitochondrial dynamics genes in the hypo-
thalamus (Figure S2C). The POMC promoter also drives cre
recombinase expression in corticotrophs and melanotrophs,
so we examined the integrity and functionality of the pituitary-
adrenal axis. Expression of pituitary genes Pomc, T box tran-
scription factor (Tpit), corticotrophin releasing hormone receptor
1 (Crhr1), pituitary-specific transcription factor 1 (Pit1), growth
hormone (Gh), and thyroid-stimulating hormone beta-chain
(Tshb) were unaltered (Figure S2D). Plasma corticosterone levels
were equivalent between controland POMCMfn2KO mice under
basal conditions, and the stress response was preserved
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 173
Figure 1. Mitochondrial Network Complexity and Mitochondria-ER Contacts in POMC Neurons Are Altered in DIO Mice
(A–C) Mitochondrial network complexity analysis in POMC neurons from 18-week-old male C57Bl/6 lean and DIO mice.
(E). Red asterisks show mitochondria in contact with ER (arrows). Note reduced number of mitochondria-ER contacts in POMC neurons from DIO mice. ER,
endoplasmic reticulum; Per, peroxisome.
(F and G) Body weight (F) and Mfn2 expression levels (G) in the hypothalamus of lean and DIO mice at different time points during a HFD time course study (n = 7–
Data are expressed as mean ± SEM. *p < 0.05. ***p < 0.001. See also Figure S1.
174 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
Figure 2. Adenoviral-Mediated Overexpression of MFN2 in the ARC of DIO Mice Ameliorates Their Metabolic Disturbances
(A) Localization studies using an Ad-GFP showing specific delivery in the mouse ARC. For comparative purposes infection in one side of the ARC is shown. 3V:
third ventricle; ME: median eminence.
(B) MFN2 western blot and densitometric quantification of ARC-enriched lysates 3 days after the delivery of Ad-Null or Ad-Mfn2. Loading control (b-Actin) is
(legend continued on next page)
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 175
(Figure S2E). Epinephrine, norepinephrine, and dopamine con-
centrations were normal in mutant mice (Figures S2F–S2H).
These results indicate that deletion of Mfn2 in pituitary cells
does not cause defects in the pituitary-adrenal axis.
Anatomical assessment of POMC neurons throughout the
tion, or somatic area (Figures S2I–S2K), indicating that Mfn2
deficiency did not alter POMC neuron differentiation and/or
Deletion of Mfn2 in POMC Neurons Causes Severe
Male POMCMfn2KO mice exhibited a marked obese phenotype
that was significant by 7 weeks of age onward (Figure 3A).
POMCMfn2KO mice were hyperphagic both under ad libitum
conditions (Figure 3B) and after a fast-refeeding test (Figure 3C),
suggesting impaired satiety mechanisms. Female mutant mice
subsequent studies in male mice.
Altered energy homeostasis in POMCMfn2KO mice was
further demonstrated by reduced oxygen consumption and
energy expenditure (Figures 3D and 3E). Given that energy
expenditure includes physical activity and thermogenesis, we
undertook studies to measure these parameters. Ambulation
activity was unchanged (control: 6,016 ± 453 counts/24 hr;
POMCMfn2KO: 7,174 ± 567 counts/24 hr; n = 7–8, NS). How-
ever, interscapular temperature adjacent to BAT was reduced
in mutant mice (Figures 3F and 3G). Consistently, mRNA expres-
sion ofBATthermogenic markers Ucp1 andperoxisome prolifer-
ator-activated receptor gamma coactivator 1-alpha (Pgc1a)
were decreased in POMCMfn2KO mice (Figure 3H). Obesity is
closely associated with alterations in glucose metabolism. As
expected, 12-week-old mutant mice exhibited hyperglycemia,
hyperinsulinemia, glucose intolerance and insulin resistance
(data not shown). Altogether, these results demonstrate that
the obese phenotype observed in POMCMfn2KO mice is
due to hyperphagia and reduced activity of the thermogenic
To assess the specificity of the metabolic consequences
caused by Mfn2 deficiency in POMC neurons, we characterized
mice lacking MFN1 (a closely related mitofusin protein) in the
same population of neurons (POMCMfn1KO). Appropriate dele-
tion of Mfn1 was assessed by the recombination event and gene
expression analysis (data not shown and Figure S3A). The pitui-
tary-adrenal axis function was preserved as suggested by
unaltered expression of pituitary genes (Figure S3B), plasma
corticosterone (Figure S3C) and catecholamines (Figures S3D–
S3F). Incontrast to POMCMfn2KOmice, energy balance param-
eters such as body weight (Figure S3G), adiposity (Figure S3H),
plasma leptin levels (Figure S3I) and feeding behavior (Figures
S3J and S3K) were unaltered in POMCMfn1KO mice. Oxygen
consumption (Figure S3L), energy expenditure (Figure S3M),
interscapular temperature (Figure S3N), and 24 hr ambulation
activity (data not shown) were comparable between mutant
homeostasis, hypothalamic neuropeptide expression was un-
changed (Figure S3O). These results demonstrate a specific
role for MFN2 in POMC neurons in the regulation of energy
Loss of MFN2 in POMC Neurons Leads to Reduced
POMC Processing into a-MSH
To investigate the mechanisms underlying the hypothalamic
phenotype of POMCMfn2KO mice, we assessed neuropeptide
gene expression. While mRNA expression of Agrp, Npy, and
Cart exhibited no differences between genotypes and appro-
priate regulation according to the nutritional status (Figure 4A),
Pomc mRNA in mutant mice was downregulated and insensitive
to the fed/fasting transition (Figure 4A). Reduced fasting Pomc
gene expression was also observed in 6-week-old mutant
mice (control: 100% ± 7% versus POMCMfn2KO: 44% ± 4%;
n = 12–16, p < 0.0001).
POMC is a prohormone that in the ARC undergoes post-
translational proteolysis by the sequential activity of different
convertases to generate a-MSH, a key melanocortin system
component implicated in energy homeostasis (Wardlaw, 2011).
extracts from 6- and 12-week-old control and POMCMfn2KO
mice. Whereas total POMC content was increased in the hypo-
thalamus from mutant mice (Figure 4B), a-MSH content was
decreased by ?50%–60% at both ages (Figure 4C). The
a-MSH/POMC ratio, a measure of POMC precursor processing,
was also decreased (Figure 4D). However, reduced a-MSH con-
centration did not seem to be the consequence of defective
mone convertase 1 (Pc1/3), prohormone convertase 2 (Pc2),
carboxypeptidase E (Cpe), a-amidating monooxygenase (Pam),
and prolylcarboxypeptidase (Prcp) were either unchanged or
regulated in opposite manner (Figures 4E and 4F).
a-MSH is released from POMC axonal terminals to the para-
ventricular nucleus (PVN) of the hypothalamus where exerts its
anorexigenic actions. Thus, we examined a-MSH peptide
expression in the PVN by immunohistochemistry and found a
dramatic reduction of a-MSH staining in neuronal projections
from mutant mice (Figures 4G and 4H).
anorectic effects of intracerebroventricular (i.c.v.) administration
of this peptide. A daily single dose of a-MSH to POMCMfn2KO
mice normalized food intake and markedly reduced body weight
gain (Figures 4I and 4J). Collectively, our data suggest that dele-
tion of Mfn2 in POMC neurons leads to reduced a-MSH concen-
tration in the hypothalamus, this defect being the cause of the
POMCMfn2KO obese phenotype.
(C–G) Body weight profile (C), adiposity (D), plasma leptin (E), food intake (F), and interscapular temperature (G).
(H) Expression of thermogenic marker Ucp-1 in BAT. Probe for Actb was used to adjust for total RNA content.
(I and J) Hypothalamic expression of neuropeptides (I) and ER stress markers (J). Probe for Hprt was used to adjust for total RNA content.
All studies were conducted in 18-week-old male lean and/or DIO mice injected intra-ARC with Ad-Null or Ad-Mfn2 (n = 10/group). Data are expressed as mean ±
SEM. *p < 0.05; **p < 0.01; ***p < 0.001.
176 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
(legend on next page)
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 177
POMCMfn2KO Mice Develop Early-Onset Leptin
The obese phenotype of POMCMfn2KO mice resulted from
were also elevated (Figure 5B), a scenario indicative of leptin
resistance. To examine leptin action in POMCMfn2KO mice we
performed a leptin sensitivity test. The anorectic effects of leptin
were clearly detectable in 12-week-old control mice. In contrast,
the effect of this hormone was blunted in POMCMfn2KO mice,
suggesting leptin resistance (Figures 5C and 5D). We next asked
whether defective leptin sensitivity in POMCMfn2KO mice was a
primary defect or secondary to the development of obesity.
Thus, we assessed leptin sensitivity at 6 weeks of age when
body weight (control: 22.1 ± 0.3 g; POMCMfn2KO: 21.0 ± 1.5 g;
n = 6, NS; Figure 3A) and adiposity (control: 13.7% ± 0.8%;
POMCMfn2KO: 14.2% ± 1.0%; n = 6, NS) were still unchanged.
reduction in food intake vehicle versus leptin: control 49% ± 3%,
POMCMfn2KO 35% ± 2%; n = 6, p = 0.005). To confirm these
findings we examined leptin signaling by double immunofluores-
cence analysis of phospho-STAT3 (pSTAT3) in POMC neurons.
Leptin administration increased pSTAT3 in POMC neurons of
control mice,but to a significantlylesserextentinPOMC neurons
from 6-week-old POMCMfn2KO mice (Figures 5G and 5H).
Together, these results indicate that deletion of Mfn2 in POMC
neurons leads to an early state of leptin resistance in the hypo-
thalamus that is independent of obesity.
Deficiency of Mfn2 in POMC Neurons Alters
Mitochondrial Morphology, Mitochondria-ER Contacts
and Induces ER Stress
It has been shown that ER stress plays a key role in the develop-
ment of leptin resistance and obesity (Ozcan et al., 2009; Zhang
et al., 2008). Given that MFN2 supports structural and functional
communication between mitochondria and ER (de Brito and
Scorrano, 2008) and that genetic loss of Mfn2 causes ER stress
(Ngoh et al., 2012; Sebastia ´n et al., 2012), we investigated
whether deletion of Mfn2 in POMC neurons leads to alterations
in mitochondria and ER homeostasis. Ultrastructure analysis by
EM showed similar density of mitochondria related to cell size
in POMC neurons from mutant mice (Figures 6A–6C), concomi-
tant with an increase in mitochondria coverage of the cell area
(Figure 6D). This change was due to enlarged mitochondria (Fig-
ure 6E), which also presented altered shape (Figure 6A, 6B, and
6F). Mitochondria from POMC neurons lacking MFN2 also con-
tained smaller and more rounded cristae (data not shown).
Remarkably, while most mitochondria in control POMC neurons
showed close interactions with ER, the number of mitochondria-
ER contacts was significantly reduced in POMC neurons from
POMCMfn2KO mice (Figures 6A, 6B, and 6G). Similar results
were observed in POMC neuron mitochondria from nonobese
6-week-old mutant mice (Figures S4A and S4B), indicating that
these alterations precede the onset of obesity. These results
imply MFN2 as a critical protein in the maintenance of mitochon-
dria morphology and mitochondria-ER interactions in hypotha-
lamic POMC neurons.
Next we conducted gene expression analysis of ER-stress
markers in the hypothalamus. By 6 weeks of age we observed
an emerging state of ER-stress in POMCMfn2KO mice, charac-
terized by the upregulation of Xbp1s and Atf4 (Figure 6H). By
12 weeks of age, the expression levels of Bip, Chop, Xbp1s,
and Atf4 were significantly increased indicating overt ER-
stress (Figure 6I). Consistently, protein expression analysis of
ER-stress markers in the hypothalamus from POMCMfn2KO
mice exhibited a similar pattern (Figures S4C and S4D). Alto-
gether, these results suggest that inactivation of Mfn2 in
POMC neurons leads to loss of mitochondria-ER interactions
and primary ER stress.
Mfn2 Deletion in POMC Neurons Alters Mitochondrial
Respiratory Capacity and Enhances ROS Production
Next, we assessed whether loss of MFN2 in POMC neurons
altered mitochondrial respiratory capacity using high-resolution
respirometry. The respiratory states Leak, oxidative phosphory-
lation due to the addition of complex I and complex I and II sub-
strates in the presence of ADP (Oxphos I and Oxphos I+II),
and maximum electron transport system capacity (ETS I+II)
were significantly reduced in hypothalamic lysates from
POMCMfn2KO mice (Figure S5A). Gene expression of Complex
I NADH dehydrogenase [ubiquinone] 1 alpha subcomplex sub-
unit 9 (Ndufa9) was reduced in the hypothalamus of mutant
results suggest impaired mitochondrial Complex I function.
Reduced activity of Complex I is associated with increased
reactive oxygen species (ROS) production. Thus, we measured
ROS levels in the hypothalamus by the assessment of protein
carbonylation, a standard marker for oxidative stress. The level
of carbonylated proteins was higher in hypothalamic lysates
from POMCMfn2KO mice (Figure S5C). Expression of ROS
detoxification enzymes catalase (Cat), glutathione peroxidase 1
(Gpx-1), and superoxide dismutase 2 (Sod2) was significantly
increased in the hypothalamus from mutant mice before and
after the onset of obesity (Figures S5D and S5E). These results
Figure 3. Mice Lacking MFN2 in POMC Neurons Are Obese due to Increased Food Intake and Reduced Thermogenesis
(A) Body weight profile on chow diet.
(B) Daily food intake in control (n = 6) and POMCMfn2KO (n = 13) mice.
(C) Fast-refeeding test in control (n = 6) and POMCMfn2KO (n = 13) mice.
(D) Daily oxygen consumption and (E) energy expenditure in control (n = 7) and POMCMfn2KO (n = 8) mice.
(F) Basal interscapular temperature adjacent to the BAT depot and (G) representative thermographic images of control (n = 6) and POMCMfn2KO (n = 7) mice.
(H) Gene expression analysis of thermogenic markers in BAT from control (n = 6) and POMCMfn2KO (n = 6) mice. Probe for Actb was used to adjust for total RNA
All studies were conducted in 12–14-week-old male control andPOMCMfn2KO mice. Data are expressed as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. See
also Figure S2.
178 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
Figure 4. Loss of MFN2 in POMC Neurons Causes Defective POMC Processing
(A) Neuropeptide expression in the hypothalamus from 12-week-old male control (n = 6–10) and POMCMfn2KO (n = 6) mice under fed and fasting conditions.
Probe for Hprt was used to adjust for total RNA content.
(B) Total hypothalamic POMC and (C) a-MSH content in 6- and 12-week-old male control (n = 6) and POMCMfn2KO (n = 7) mice.
(D) POMC processing as measured by a-MSH/POMC ratio.
(E–F) Expression analysis of POMC processing genes in the hypothalamus from 6- and 12-week-old male control (n = 9–13) and POMCMfn2KO (n =
(legend continued on next page)
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 179
might reflect a compensatory mechanism aimed to reduce the
excess of ROS generated.
Our data show specific mitochondrial function disturbances
and enhanced ROS production in the hypothalamus of
POMCMfn2KO mice. Increased ROS levels in the hypothalamus
causes reduced food intake and body weight gain in rodents
(Diano et al., 2011; Gyengesi et al., 2012), so it is unlikely that
type of POMCMfn2KO mice. Nevertheless, we undertook exper-
iments to assess this hypothesis by i.c.v. injection of the ROS
scavenger honokiol. Consistent with previous results (Diano
et al., 2011), central honokiol administration significantly
increased food intake and body weight gain in control mice,
and these effects were further magnified in mutant mice (Figures
S5F and S6G), indicating that enhanced ROS generation in the
hypothalamus is not the cause of the obesogenic phenotype of
Inhibition of Hypothalamic ER Stress by Chemical
Chaperone Treatment Reverses the Obesogenic
Phenotype of POMCMfn2KO Mice
Chemical chaperones, such as 4-phenyl butyric acid (4-PBA) or
tauroursodeoxycholic acid (TUDCA), are able to relieve ER
stress and improve ER function in different tissues including
the hypothalamus (Ozcan et al., 2009; Zhang et al., 2008). To
test whether the metabolic defects observed in POMCMfn2KO
mice resulted from hypothalamic ER stress, we conducted
i.c.v. injections of either 4-PBA or TUDCA to control and
POMCMfn2KO mice. 4-PBA administration did not change
food intake or body weight gain in control mice (Figures 7A
and 7B). In contrast, this treatment completely normalized food
intake and body weight in mutant mice (Figures 7A and 7B).
Consistent with this, adiposity (Figure 7C) and plasma leptin
levels (Figure 7D) were restored in POMCMfn2KO mice treated
with 4-PBA. Chaperone treatment also normalized the expres-
sion of Xbp1s, Chop, Atf4, and PERK phosphorylation in the
hypothalamus of mutant mice, indicating ER stress relieve (Fig-
ure 7E and data not shown). Remarkably, 4-PBA treatment
normalized Pomc mRNA expression as well as the content of
POMC precursor, a-MSH, and a-MSH/POMC ratio in the hypo-
thalamus from POMCMfn2KO mice (Figures 7F–7I). a-MSH
immunoreactivity in the PVN was also regained in mutant mice
after 4-PBA administration (Figures 7J and 7K). The beneficial
effects of 4-PBA were not associated with reduced hypothalam-
ic levels of carbonylated proteins (Figure S5H), demonstrating
that the normalization of the phenotype was not the conse-
quence of changes in hypothalamic ROS. Central TUDCA treat-
mentessentially reproduced theeffects of 4-PBAbysignificantly
improving the metabolic alterations of POMCMfn2KO, although
it was unable to completely restore their phenotype (Figures
S6A–S6K). Collectively, these results indicate that i.c.v. treat-
ment with two structurally unrelated chemical chaperones is
able to recover the physiological andmolecular alterations found
in POMCMfn2KO mice, suggesting that these are the conse-
quence of hypothalamic ER stress.
Here, weshow thatMFN2 in POMC neurons isan essential regu-
lator of whole-body energy balance. We suggest a model
whereby disruption of proper mitochondria-ER homeostasis
through Mfn2 deletion in POMC neurons leads to primary ER
stress, defective a-MSH processing, leptin resistance, and
obesity. Thus, our data support the notion that MFN2 is a molec-
ular determinant implicated in the development of ER stress-
nutrient fluctuations and integrate coordinated responses aimed
to maintain cell homeostasis (Baltzer et al., 2010; Youle and van
der Bliek, 2012). This is achieved through the establishment of
contact sites between both organelles that participate in essen-
tial bidirectional communication processes (Rowland and
Voeltz, 2012). Given that mitochondrial morphology is continu-
ously changed through fusion and fission events, a tight coordi-
nation between mitochondrial dynamics and interorganelle
interactions is crucial. MFN2 has been extensively implicated
in mitochondrial fusion (Zorzano et al., 2010), but recent data
also demonstrate a key role for MFN2 tethering ER with mito-
chondria (de Brito and Scorrano, 2008). This dual role of MFN2
is unique and makes it a good candidate to coordinate ER and
mitochondrial functions, especially those modulated by meta-
bolic stress. In agreement with this, we found that nutrient
excess in DIO mice altered the mitochondrial
complexity and reduced the number of mitochondria-ER con-
tacts in POMC neurons. This was correlated with specific
downregulation of Mfn2 expression in the hypothalamus. Over-
expression of MFN2 in the ARC of DIO mice ameliorated their
metabolic defects, suggesting a causative role for MFN2 in the
development ofmetabolicalterations under nutrientexcesscon-
ditions. In contrast, high-fat feeding stimulated a mitochondrial
fusion-like process in AgRP neurons (Dietrich et al., 2013 [this
issue of Cell]) suggesting cell-type-specific roles for MFN2
upon mitochondrial dynamics. Collectively, this data support
the current idea that mitochondrial fusion is regulated by energy
demand and stress (Youle and van der Bliek, 2012) and suggest
that MFN2 is a key determinant of a molecular pathway that
integrates nutritional changes with the mitochondria-ER axis to
adapt and maximize the function of this system.
Consistent with a key role for MFN2 in metabolic regulation,
deletion of Mfn2 in POMC neurons caused a dramatic obese
phenotype characterized by increased adiposity, hyperphagia,
and reduced energy expenditure. ARC POMC neurons are major
targets of leptin action, stimulating Pomc transcription through
the JAK2-STAT3 pathway and triggering the release of the
anorexigenic neuropeptide a-MSH from POMC axon terminals
(Kitamura et al., 2006; Mu ¨nzberg et al., 2003). In line with this,
POMCMfn2KO mice exhibited attenuated leptin effects and
(G) Representative immunofluorescence images showing a-MSH staining in the PVN from 12-week-old male control (n = 3) and POMCMfn2KO (n = 3) mice and
(H) integrated density quantification.
(I) Food intake and (J) body weight gain in 12-week-old male control (n = 4) and POMCMfn2KO (n = 5) mice after acute i.c.v. injection of a-MSH.
Data are expressed as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. 3V: third ventricle. PVN: paraventricular nucleus. See also Figure S3.
180 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
(legend on next page)
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 181
reduced pSTAT3 staining in POMC neurons before the onset of
obesity. In addition, Pomc mRNA expression was reduced and
unresponsive to the fed/fasting transition in POMCMfn2KO
rons cause primary leptin resistance independent of obesity.
Hypothalamic ER stress has recently emerged as a causal
factor in the development of central leptin resistance (Hosoi
et al., 2008; Ozcan et al., 2009; Won et al., 2009; Zhang et al.,
2008). Given the cellular localization and function of MFN2, we
reasoned that this protein could represent a molecular link
between hypothalamic ER stress and leptin resistance. Ultra-
structural analysis of POMC neurons lacking MFN2 revealed
spherical and enlarged mitochondria, a morphology that is in
line with previous studies (Chen et al., 2007; Lee et al., 2012;
Papanicolaou et al., 2011). Remarkably, we observed a reduc-
tion in mitochondria-ER contacts in POMC neurons from
POMCMfn2KO obese mice. The extension of this loss of
contact sites was strikingly similar to those observed in DIO
mice, suggesting that defective mitochondria-ER contacts
may underlie leptin resistance and obesity development.
Furthermore, hypothalamic gene and protein expression anal-
ysis showed incipient upregulation of ER stress markers in non-
obese POMCMfn2KO mice, a defect that was further enhanced
in adult mutant mice. Deletion of Mfn2 in AgRP neurons did not
lead to alterations in mitochondria-ER contacts (Dietrich et al.,
2013), further indicating divergent roles for MFN2 in a cell-
The neurobiological cause of hypothalamic ER stress-induced
leptin resistance is still unknown but likely due to a combination
of defects. For example, in vitro and in vivo data indicate that ER
stress and the activation of the UPR signaling network directly
block leptin signaling by reducing pSTAT3 levels (Ozcan et al.,
2009). Consistently, POMCMfn2KO mice displayed reduced
pSTAT3 staining in POMC neurons. Furthermore, the ER is
involved in the synthesis, folding and transport of secretory pro-
teins (Rowland and Voeltz, 2012). It is plausible that dysfunc-
tional ER may interfere with proper synthesis and release of
key neuropeptides that mediate the anorectic effects of leptin,
thereby altering whole-body energy balance. Hence, ER stress
in POMC neurons could result in defective POMC processing
and/or a-MSH production. We observed that, although POMC
protein was increased likely as a consequence of compensatory
mechanisms, total a-MSH content and processing rate was
reduced in mutant mice. POMC processing alterations were
not due to reduced expression of convertases. In fact, upregula-
tion of some of these convertases and downregulation of a-MSH
degrading enzyme Prcp (Wallingford et al., 2009) was observed,
probably as a molecular adaptation to maintain the appropriate
anorexigenic tone. Defective POMC processing was observed
before the onset of obesity suggesting that this chain of events
precede, and are likely the cause, of the development of obesity
in POMCMfn2KO mice.
We further demonstrated that ER stress was the underlying
cause of the metabolic phenotype of POMCMfn2KO mice
through a rescue experiment. I.c.v. administration of two struc-
turally unrelated chemical chaperones, such as 4-PBA or
TUDCA, was able to fully recover or improve to great extent
the physiological and molecular alterations of POMCMfn2KO
mice. These results support the notion that loss of MFN2 in
POMC neurons cause abnormal hypothalamic ER homeostasis
and function, leading to defective POMC neuropeptide produc-
tion and processing. Our data are consistent with reports indi-
cating that deletion of Mfn2 causes ER stress both in vitro and
in vivo (Ngoh et al., 2012; Sebastia ´n et al., 2012).
Recent studies suggest that neuronal Mfn2 deficiency or dele-
althoughdirect assessment ofdefective axonal growthislacking
in some of these reports (Chen et al., 2007; Lee et al., 2012;
Misko et al., 2012; Pham et al., 2012). POMCMfn2KO mice ex-
hibited a dramatic reduction in a-MSH in neuronal projections
to the PVN without alterations in POMC neuron number. This
could be interpreted as evidence supporting compromised
axon integrity, but the fact that 4-PBA and TUDCA treatment
was able to restore a-MSH immunoreactivity in the PVN argues
against this hypothesis. Similarly, loss of MFN2 in AgRP neurons
did not cause axon degeneration (Dietrich et al., 2013). Collec-
tively, these evidences suggest cell-specific functions of MFN2
in relation to neuronal survival and axon integrity.
MFN1 and MFN2 are highly homologous proteins, but a
number of studies using conditional mouse mutants have shown
that they are not entirely redundant (Chen et al., 2007; Lee et al.,
et al., 2012). Our data further demonstrate specific and nonover-
lapping functions for MFN1 and MFN2, as deletion of Mfn1 in
POMC neurons did not cause alterations in energy balance. It
is now generally recognized that MFN1 is mainly implicated in
mitochondrial fusion, whereas MFN2 could mediate additional
functions such as mitochondria-ER tethering and mitochondrial
transport. Our data indicate a main role for MFN2 in POMC
neurons upon ER homeostasis maintenance.
Defective ER and mitochondrial function has been extensively
associated with metabolic disorders. In previous studies, we
have demonstrated the significance of MFN2 in skeletal muscle
and liver in the pathophysiology of obesity and type 2 diabetes in
both humans and rodents (Bach et al., 2005; Bach et al., 2003;
Herna ´ndez-Alvarez et al., 2009; Herna ´ndez-Alvarez et al.,
2010; Sebastia ´n et al., 2012). Here we further expanded this
knowledge to hypothalamic POMC neurons, a key population
of neurons implicated in whole-body energy homeostasis, and
provided evidence for a causative link between nutrient excess
Figure 5. Deficiency of Mfn2 in POMC Neurons Leads to Primary Leptin Resistance
(A) Body fat content and (B) plasma leptin in male control (n = 6) and POMCMfn2KO (n = 7) mice at 12–14 weeks of age.
(C and D) Body weight gain and cumulative food intake in 12–14-week-old male control (n = 7) and POMCMfn2KO (n = 8) mice after 3 day leptin injection.
(E and F) Daily body weight gain and food intake in 6-week-old male control (n = 6) and POMCMfn2KO (n = 6) mice after an acute leptin sensitivity test.
(G and H) Representative images showing double immunofluorescence examining pSTAT3 in POMC neurons and percentage of colocalization after vehicle or
leptin stimulation in 6-week-old POMCZ/EG (n = 3) and POMCMfn2KOZEG (n = 3) mice.
Data are expressed as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. 3V: third ventricle. ME: median eminence. ns: not significant. See also Figure S4.
182 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
Figure 6. Deletion of Mfn2 in POMC Neurons Alters Mitochondrial Morphology, Mitochondrial-ER Contacts and Induces ER Stress
(Aand B)RepresentativeelectronmicroscopyimagesofPOMCneuronsfrom12-week-oldmalecontrol (A)andPOMCMfn2KO(B)mice.Mitochondria (Mito:pink
areas) and endoplasmic reticulum (ER: yellow areas) are shown. Red asterisk shows mitochondria in contact with ER.
(C–F) Mitochondria density (C), mitochondria coverage (D), mitochondria area (E), and mitochondria aspect ratio (AR) (F) in POMC neurons from 12-week-old
male control (n = 5; 1,198 mitochondria from 30 POMC neurons) and POMCMfn2KO (n = 3; 530 mitochondria from 15 POMC neurons) mice.
(G) Percentage of mitochondria-ER contacts in POMC neurons from 12-week old male control (n = 5; 32 POMC neurons) and POMCMfn2KO (n = 3; 17 POMC
(H) Gene expression analysis of ER-stress markers in the hypothalamus from 6-week-old male control (n = 10–15) and POMCMfn2KO (n = 8–12) mice.
(I) Geneexpression analysis of ER-stress markers inthe hypothalamus from 12-week-old male control(n = 21–23) andPOMCMfn2KO (n = 16–18) mice. Probefor
Hprt was used to adjust for total RNA content.
Data are expressed as mean ± SEM. *p < 0.05; ***p < 0.001. See also Figure S5.
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 183
(legend on next page)
184 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.
and alterations in mitochondrial dynamics, mitochondria-ER
interactions, and Mfn2 expression. Importantly, our data also
established MFN2 in POMC neurons as an essential regulator
of systemic energy balance by fine-tuning the mitochondrial-
ER axis homeostasis and function. This previously unrecognized
role for MFN2 argues for a critical involvement in mediating ER
stress-induced leptin resistance.
Mice and Diets
C57BL/6 mice were purchased from Harlan Europe. The generation of POMC-
Cre mice has been previously reported (Xu et al., 2005). Mfn1loxp/loxpand
Mfn2loxp/loxpmice were provided by Dr. David C. Chan (Chen et al., 2007)
through the Mutant Mouse Regional Resource Center. To generate POMC-
specific Mfn1 and Mfn2 knockout mice (POMCMfn1KO and POMCMfn2KO,
respectively), POMC-Cre mice were crossed with Mfn1loxp/loxpor Mfn2loxp/loxp
mice. Colonies were maintained by breeding POMC-cre; loxp/loxp mice with
loxp/loxp mice. To generate mice expressing GFP in cells harboring the dele-
tion event, mice were intercrossed with Z/EG mice (Novak et al., 2000). Mice
were maintained on a 12:12 hr light-dark cycle with free access to water and
standard chow (Harlan Research Laboratories) or HFD (45% Kcal fat;
Research Diets) for 12 weeks (starting at 6 weeks of age). In vivo studies
were performed with approval of the University of Barcelona Ethics Com-
mittee, complying with current Spanish and European legislation.
Body weights were determined weekly. Blood samples were collected via tail
vein or trunk bleeds using a capillary collection system (Sarstedt). Blood
glucose was measured using a Glucometer (Arkray). Plasma insulin and leptin
were analyzed by ELISA (Crystal Chem) and corticosterone by enzyme
immunoassay (IDS). Catecholamines were measured by ELISA (Labor
Diagnostika Nord). Glucose tolerance tests were performed on overnight
fasted mice. D-glucose (2 g/kg) was injected intraperitoneally (i.p.) and blood
glucose determined at the indicated time points. Insulin sensitivity tests were
performed on 4 hr food-deprived mice. Insulin (0.4 IU/kg) was injected i.p. and
blood glucose determined at the indicated time points.
Body Composition, Thermal Imaging and Indirect Calorimetry
Whole-body composition was measured using NMR imaging (EchoMRI). Heat
production was visualized using a high-resolution infrared camera (FLIR
Systems) as previously described (Czyzyk et al., 2012). Indirect calorimetry
was assessed using a TSE LabMaster modular research platform (TSE
Systems) as previously described (Czyzyk et al., 2012).
Feeding Studies and Leptin Sensitivity Tests
Mice were singly housed and acclimatized for 1 week prior to study. Food
intake was measured for 5 consecutive days. For fast-refeeding studies,
mice were overnight fasted and refed with a preweighed amount of food.
Food intake was measured at the indicated time points. For leptin sensitivity
tests, mice were acclimatized by subjecting them to handling and sham injec-
tions. Leptin tests were conducted in a crossover fashion. Twelve-week-old
control and POMCMfn2KO mice were i.p. injected with either 1.5 mg/g of
mouse leptin (R&D Systems) or vehicle twice a day (1 hr before lights out
(7 p.m.) and at 8.00 a.m.) for 3 consecutive days. Six-week-old control and
POMCMfn2KO mice were submitted to an acute leptin test to assure similar
body weights during the protocol. I.p. administration of either 5 mg/g of leptin
(R&D Systems) or vehicle 1 hr before lights-out was performed. Food intake
and body weights were recorded daily.
Hypothalami and pituitaries were harvested and immediately frozen in liquid
nitrogen. mRNA was isolated using Trizol (Invitrogen). Retrotranscription and
quantitative RT-PCR (qPCR) was performed as previously described (Claret
et al., 2007). Proprietary Taqman Gene Expression assay FAM/TAMRA
primers used (Applied Biosystems) are detailed in Extended Experimental
Electron Microscopy and Mitochondrial Analysis
Electron microscopy and analysis was performed as described (Dietrich et al.,
2013). Detailed protocols are provided in Extended Experimental Procedures.
Hypothalamic POMC and a-MSH Protein Content
Hypothalami were sonicated in 500 ml of 0.1N HCl solution. Lysates were
centrifuged and supernatants used for the quantification of POMC (Uscn Life
Science) and a-MSH (Phoenix Pharmaceuticals) by ELISA. Protein concentra-
tion was determined by Bradford.
Brains were processed and immunohistochemistry conducted as described
(Claret et al., 2007; Claret et al., 2011). Antibodies used are detailed in
Extended Experimental Procedures. Imaging was performed using a Leica
DMI 6000B microscope. a-MSH integrated density after image skeletonization
was calculated using ImageJ software.
POMC Neuron Count and Area
The distribution and number of POMC neurons were determined as described
(Claret et al., 2007). Average somatic area was analyzed in > 500 POMC
neurons (n = 3 mice/genotype). The area occupied by POMC neurons was
manually scored using ImageJ software.
Quantification of Leptin-Induced pSTAT3 in POMC Neurons
Six-week-old male POMCZ/EG and POMCMfn2KOZ/EG mice were injected
i.p. with either saline or 5 mg/g mouse leptin (R&D Systems). Mice were trans-
cardially perfused 45 min after injection, and brains were processed for
pSTAT3 immunohistochemistry (1:1,000, Cell Signaling). Double GFP and
pSTAT3 labeling was determined in matched sections (n = 3 mice/group).
Mitochondrial Network Complexity Analysis in POMC Neurons
Mitochondrial morphology and network complexity in POMC neurons was
assessed by confocal microscopy and computer-assisted analysis. Detailed
methodology is provided in Extended Experimental Procedures.
I.c.v. Cannulation and Treatments
I.c.v. surgery was performed as previously described (Al-Qassab et al., 2009).
Detailed protocols and treatments are provided in Extended Experimental
Data are expressed as mean ± SEM. p Values were calculated using unpaired
Student’s t test or one-way ANOVA with post hoc Tukey tests as appropriate.
p % 0.05 was considered significant.
Figure 7. Restoration of Energy Homeostasis in POMCMfn2KO Mice by 4-PBA Administration
Adiposity; (D) Plasma leptin; (E) Expression of ER stress markers in the hypothalamus (n = 4/genotype/treatment); (F) Pomc mRNA expression (n = 4/genotype/
treatment); (G) Hypothalamic POMC content (n= 4/genotype/treatment); (H) Hypothalamic a-MSH content (n = 4/genotype/treatment); (I)a-MSH/POMCratio;(J)
Representative immunofluorescence images showing a-MSH staining in the PVN and (K) integrated density quantification (n = 3/genotype/treatment). Probe for
Hprt was used to adjust for total RNA content. Data are expressed as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. eWAT: epididymal white adipose tissue;
BW: body weight; V: vehicle. 3V: third ventricle. PVN: paraventricular nucleus. See also Figure S6.
Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc. 185
SUPPLEMENTAL INFORMATION Download full-text
Supplemental Information includes Extended Experimental Procedures and
six figures and can be found with this article online at http://dx.doi.org/10.
We are grateful to Dr Gregory S Barsh (Stanford University) for providing
Pomc-cre mice, Dr Maria Calvo and Anna Bosch from the Confocal micro-
scopy unit (SCT-UB), Merce ` Monfar (UPV-CBATEG), and members from the
Diabetes and Obesity Lab. This work has been supported by: RecerCaixa
Grant 2010ACUP_00275 (M.C.); EFSD/Lilly Fellowship Award (M.C.); Minis-
terio de Ciencia e Innovacio ´n (MICINN), Instituto de Salud Carlos III (ISCIII)
Grant PI10/01074 (M.C.); MICINN Grant SAF2010-19527 (R.G.); Ministerio
de Economia y Competitividad (MINECO) Grant BFU2012-35255 (RN); Xunta
de Galicia Grant EM2012/039 and 2012-CP069 (R.N.); CIBERobn (R.N.); Euro-
pean Community’s Seventh Framework Programme Grant ERC-2011-StG-
OBESITY53-281408 (R.N.); MINECO Grant SAF2008-03803 (A.Z.); Generalitat
de Catalunya Grant 2009SGR915 (A.Z.); CIBERDEM (A.Z.); INTERREG IV-
B-SUDOE-FEDER (DIOMED, SOE1/P1/E178) (A.Z.); European Commission
Seventh Framework Programme (FP7), MITIN Grant HEALTH-F4-2008-
223450 (A.Z.); MICINN Grant BFU2011- 24679 (P.M.G.-R.); ISCIII Grant
PI10/00290 (A.B.). T.L.H. and M.O.D. are supported by NIH (DP1DK006850,
R01AG040236, P01NS062686), American Diabetes Association, Helmholtz
society (ICEMED) and by Conselho Nacional de Desenvolvimento Cientı ´fico
e Tecnolo ´gico (401476/2012-0, Brazil). M.S. is a recipient of an undergraduate
grant from the University of Barcelona. M.C. is a recipient of a Miguel Servet
contract (CP09/00233) from MICINN-ISCIII. P.M.G.-R. is a recipient of a
Ramon y Cajal contract (RYC-2009-05158) from MICINN. Some of these
grants are cofinanced by the European Regional Development Fund ‘‘A way
to build Europe.’’ This work was carried out in part at the Esther Koplowitz
Received: February 20, 2013
Revised: July 9, 2013
Accepted: August 22, 2013
Published: September 26, 2013
Al-Qassab, H., Smith, M.A., Irvine, E.E., Guillermet-Guibert, J., Claret, M.,
Choudhury, A.I., Selman, C., Piipari, K., Clements, M., Lingard, S., et al.
(2009). Dominant role of the p110beta isoform of PI3K over p110alpha in
energy homeostasis regulation by POMC and AgRP neurons. Cell Metab.
Bach, D., Pich, S., Soriano, F.X., Vega, N., Baumgartner, B., Oriola, J., Dau-
gaard, J.R., Lloberas, J., Camps, M., Zierath, J.R., et al. (2003). Mitofusin-2
determines mitochondrial network architecture and mitochondrial meta-
bolism. A novel regulatory mechanism altered in obesity. J. Biol. Chem. 278,
Bach, D., Naon, D., Pich, S., Soriano, F.X., Vega, N., Rieusset, J., Laville, M.,
Guillet, C., Boirie, Y., Wallberg-Henriksson, H., et al. (2005). Expression of
Mfn2, the Charcot-Marie-Tooth neuropathy type 2A gene, in human skeletal
muscle: effects of type 2 diabetes, obesity, weight loss, and the regulatory
role of tumor necrosis factor alpha and interleukin-6. Diabetes 54, 2685–2693.
Baltzer, C., Tiefenbo ¨ck, S.K., and Frei, C. (2010). Mitochondria in response to
nutrients and nutrient-sensitive pathways. Mitochondrion 10, 589–597.
Bjørbaek, C. (2009). Central leptin receptor action and resistance in obesity.
J. Investig. Med. 57, 789–794.
Chen, H., Detmer, S.A., Ewald, A.J., Griffin, E.E., Fraser, S.E., and Chan, D.C.
(2003). Mitofusins Mfn1 and Mfn2 coordinately regulate mitochondrial fusion
and are essential for embryonic development. J. Cell Biol. 160, 189–200.
Chen, H., McCaffery, J.M., and Chan, D.C. (2007). Mitochondrial fusion
protects against neurodegeneration in the cerebellum. Cell 130, 548–562.
Claret, M., Smith, M.A., Batterham, R.L., Selman, C., Choudhury, A.I., Fryer,
L.G., Clements, M., Al-Qassab, H., Heffron, H., Xu, A.W., et al. (2007). AMPK
is essential for energy homeostasis regulation and glucose sensing by
POMC and AgRP neurons. J. Clin. Invest. 117, 2325–2336.
Claret, M., Smith, M.A., Knauf, C., Al-Qassab, H., Woods, A., Heslegrave, A.,
Piipari, K., Emmanuel, J.J., Colom, A., Valet, P., et al. (2011). Deletion of Lkb1
in pro-opiomelanocortin neurons impairs peripheral glucose homeostasis in
mice. Diabetes 60, 735–745.
Czyzyk, T.A., Romero-Pico ´, A., Pintar, J., McKinzie, J.H., Tscho ¨p, M.H.,
Statnick, M.A., and Nogueiras, R. (2012). Mice lacking d-opioid receptors
resist the development of diet-induced obesity. FASEB J. 26, 3483–3492.
de Brito, O.M., and Scorrano, L. (2008). Mitofusin 2 tethers endoplasmic
reticulum to mitochondria. Nature 456, 605–610.
Diano, S., Liu, Z.W., Jeong, J.K., Dietrich, M.O., Ruan, H.B., Kim, E., Suyama,
S., Kelly, K., Gyengesi, E., Arbiser, J.L., et al. (2011). Peroxisome proliferation-
associated control of reactive oxygen species sets melanocortin tone and
feeding in diet-induced obesity. Nat. Med. 17, 1121–1127.
Die ´guez, C., Vazquez, M.J., Romero, A., Lo ´pez, M., and Nogueiras, R. (2011).
Hypothalamic control of lipid metabolism: focus on leptin, ghrelin and melano-
cortins. Neuroendocrinology 94, 1–11.
Dietrich, M., Liu, Z.-W., and Horvath, T.L. (2013). Mitochondrial dynamics
controlled by mitofusins regulate Agrp neuronal activity and diet-induced
obesity. Cell 155, this issue, 188–199.
Gamber, K.M., Huo, L., Ha, S., Hairston, J.E., Greeley, S., and Bjørbæk, C.
creases susceptibility to diet-induced obesity. PLoS ONE 7, e30485.
Gyengesi, E., Paxinos, G., and Andrews, Z.B. (2012). Oxidative Stress in the
Hypothalamus: the Importance of Calcium Signaling and Mitochondrial ROS
in Body Weight Regulation. Curr. Neuropharmacol. 10, 344–353.
Herna ´ndez-Alvarez, M.I., Chiellini, C., Manco, M., Naon, D., Liesa, M., Palacı ´n,
M., Mingrone, G., and Zorzano, A. (2009). Genes involved in mitochondrial
biogenesis/function are induced in response to bilio-pancreatic diversion in
morbidly obese individuals with normal glucose tolerance but not in type 2
diabetic patients. Diabetologia 52, 1618–1627.
Herna ´ndez-Alvarez, M.I., Thabit, H., Burns, N., Shah, S., Brema, I., Hatunic,
M., Finucane, F., Liesa, M., Chiellini, C., Naon, D., et al. (2010). Subjects
with early-onset type 2 diabetes show defective activation of the skeletal
muscle PGC-1alpha/Mitofusin-2 regulatory pathway in response to physical
activity. Diabetes Care 33, 645–651.
Hosoi, T., Sasaki, M., Miyahara, T., Hashimoto, C., Matsuo, S., Yoshii, M., and
Ozawa, K. (2008). Endoplasmic reticulum stress induces leptin resistance.
Mol. Pharmacol. 74, 1610–1619.
Kitamura, T., Feng, Y., Kitamura, Y.I., Chua, S.C., Jr., Xu, A.W., Barsh, G.S.,
Rossetti, L., and Accili, D. (2006). Forkhead protein FoxO1 mediates Agrp-
dependent effects of leptin on food intake. Nat. Med. 12, 534–540.
Lee, S., Sterky, F.H., Mourier, A., Terzioglu, M., Cullheim, S., Olson, L., and
Larsson, N.G. (2012). Mitofusin 2 is necessary for striatal axonal projections
of midbrain dopamine neurons. Hum. Mol. Genet. 21, 4827–4835.
Mandl, J., Me ´sza ´ros, T., Ba ´nhegyi, G., Hunyady, L., and Csala, M. (2009).
Endoplasmic reticulum: nutrient sensor in physiology and pathology. Trends
Endocrinol. Metab. 20, 194–201.
Misko, A.L., Sasaki, Y., Tuck, E., Milbrandt, J., and Baloh, R.H. (2012). Mitofu-
sin2 mutations disrupt axonal mitochondrial positioning and promote axon
degeneration. J. Neurosci. 32, 4145–4155.
Mu ¨nzberg, H., Huo, L., Nillni, E.A., Hollenberg, A.N., and Bjørbaek, C. (2003).
Role of signal transducer and activator of transcription 3 in regulation of
hypothalamic proopiomelanocortin gene expression by leptin. Endocrinology
Myers, M.G., Jr., and Olson, D.P. (2012). Central nervous system control of
metabolism. Nature 491, 357–363.
Ngoh, G.A., Papanicolaou, K.N., and Walsh, K. (2012). Loss of mitofusin 2
promotes endoplasmic reticulum stress. J. Biol. Chem. 287, 20321–20332.
186 Cell 155, 172–187, September 26, 2013 ª2013 Elsevier Inc.