Acquirement of Brown Fat Cell Features by Human White Adipocytes*

Obesity, i.e. an excess of white adipose tissue (WAT), predisposes to the development of type 2 diabetes and cardiovascular disease. Brown adipose tissue is present in rodents but not in adult humans. It expresses uncoupling protein 1 (UCP1) that allows dissipation of energy as heat. Peroxisome proliferator-activated receptor γ (PPARγ) and PPARγ coactivator 1α (PGC-1α) activate mouse UCP1 gene transcription. We show here that human PGC-1α induced the activation of the human UCP1 promoter by PPARγ. Adenovirus-mediated expression of human PGC-1α increased the expression of UCP1, respiratory chain proteins, and fatty acid oxidation enzymes in human subcutaneous white adipocytes. Changes in the expression of other genes were also consistent with brown adipocyte mRNA expression profile. PGC-1α increased the palmitate oxidation rate by fat cells. Human white adipocytes can therefore acquire typical features of brown fat cells. The PPARγ agonist rosiglitazone potentiated the effect of PGC-1α on UCP1 expression and fatty acid oxidation. Hence, PGC-1α is able to direct human WAT PPARγ toward a transcriptional program linked to energy dissipation. However, the response of typical white adipocyte targets to rosiglitazone treatment was not altered by PGC-1α. UCP1 mRNA induction was shown in vivo by injection of the PGC-1α adenovirus in mouse white fat. Alteration of energy balance through an increased utilization of fat in WAT may be a conceivable strategy for the treatment of obesity.

predisposes to the development of an array of metabolic disturbances leading to type 2 diabetes and cardiovascular disease. Although it shares many features with WAT, brown adipose tissue (BAT) is specialized in adaptive thermogenesis (1). Differences in gene expression between WAT and BAT, most notably at the mitochondrial level, explain the thermogenic capacity of BAT. Fatty acid oxidation enzymes and respiratory chain components are highly expressed in BAT contributing to a high oxidative capacity. The activity of ATP synthase is low because of a defect in expression of the P1 gene (2). However, the most distinguishing feature of BAT is the expression of uncoupling protein 1 (UCP1) (3). UCP1 is a 32-kDa protein expressed in the inner membrane of the mitochondria. UCP1 allows the dissipation of the proton electrochemical gradient generated by the respiratory chain. Uncoupling between oxygen consumption and ATP synthesis promotes energy dissipation as heat. The mechanism of action of UCP1 is still controversial. One model depicts UCP1 as a true proton transporter, whereas another model states that UCP1 catalyzes a fatty acid protonophoretic cycle (4). Fatty acids and retinoids have been shown to activate UCP1 (5,6). In neonatal mammals, hibernators and rodents, cold-induced thermogenesis in BAT contributes to the maintenance of body temperature. Fuel is provided as fatty acids derived from BAT and WAT lipolysis. In rodents, BAT also participates in diet-induced thermogenesis and may thereby control the energy efficiency of food (7). UCP1 biosynthesis is mainly controlled at the transcriptional level. During cold exposure, sympathetic nervous system stimulation of BAT is the primary signal that activates UCP1 gene expression. Retinoic acid (RA) and thyroid hormones are other positive regulators. A critical enhancer has been characterized in rodent UCP1 genes (8,9). This region is required for catecholamine and RA stimulation. The enhancer contains a peroxisome proliferator-activated receptor ␥ (PPAR␥) responsive element that mediates the stimulation induced by thiazolidinediones (TZD) (10). In cooperation with PPAR␥, the PPAR␥ coactivator PGC-1␣, has been shown to induce mouse UCP1 gene transcription (11). It also stimulates the expression of electron transport chain genes and mitochondrial biogenesis, through induction of nuclear respiratory factors 1 and 2 (12). PGC-1␣ expression is increased in response to cold exposure and ␤-adrenergic stimulation (11).
BAT is present throughout the life in rodents but disappears soon after birth in large mammals. In human fetus and newborn children, it is found in the cervical, axillary, perirenal, and periadrenal depots (13). There are no BAT depots in adults, and UCP1 mRNA is expressed at very low levels in WAT (14). BAT is not thought to contribute to a significant part of thermogenesis (15). However, UCP1 is expressed in hibernomas and in perirenal WAT of adult patients with phaeochromocytoma and primary aldosteronism revealing that UCP1 expression can be induced in rare tumors and endocrinological disorders (16,17). Pharmacotherapy targeted at molecular pathways that regulate adaptive thermogenesis provides a plausible and safe means of increasing energy expenditure (18). Reactivation of brown adipocytes is therefore an important goal. Studies on human white adipocytes are mandatory to substantiate the proof of concept. In an attempt to promote a metabolic shift in white fat cells from lipid storage toward fatty acid utilization, human subcutaneous white adipocytes were transduced with an adenovirus expressing PGC-1␣. The cells acquire features of brown adipocytes, i.e. an induction of UCP1 and respiratory chain gene expression, and an increased capacity to oxidize fatty acid. Conversion of white into brown adipocytes may therefore constitute a strategy to regulate fat mass in humans.

Adenoviral Expression System and Adenofection Experiment in CV-1
Cells-Recombinant adenovirus was generated as described (19). The full-length human PGC-1␣ cDNA (20) was cloned into the pAdEasy parent plasmid. Recombination between the pAdEasy and pAdTrack vectors and production of the PGC-1␣ adenovirus was performed at the Laboratoire de Thérapie Génique de Nantes. The virus contains, in tandem, the green fluorescent protein (GFP) gene and the PGC-1␣ cDNA downstream of separate cytomegalovirus promoters. An adenovirus containing only the GFP gene was used as control. Viral titers were, respectively, 1.7 ϫ 10 11 and 1.4 ϫ 10 11 infectious particles per ml. Adenofection experiments were performed in CV-1 cells (ATCC, Manassas, VA) cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (Invitrogen, Cergy Pontoise, France). The 6.3-kb UCP1 promoter-chloramphenicol acetyltransferase gene construct (21) was cotransfected with expression vectors for PPAR␥2 (from Bruce Spiegelman, Dana-Farber Cancer Institute, Boston, MA) and retinoic acid X receptor ␣ (RXR␣) (from Pierre Chambon, IGBMC, Strasbourg, France) and a cytomegalovirus promoter-␤-galactosidase gene vector to check for transfection efficiency. The PGC-1␣ or the GFP adenoviruses were added to the LipofectAMINE (Invitrogen) transfection mixture at a multiplicity of infection (m.o.i.) of 200. Cells were exposed for 6 h to the transfection mixture. Chloramphenicol acetyltransferase activity was assayed on cell extracts 72 h post-adenofection.
Differentiation of Human Preadipocytes, Adenovirus Infection, and Flow Cytometry-Subcutaneous abdominal adipose tissue was obtained from female subjects undergoing plastic surgery in agreement with French laws on biomedical research. Human adipocytes in primary culture were differentiated as described by Hauner et al. (22) with modifications (23). Stromal cells prepared from WAT were cultured for 13 days in a chemically defined medium. At day 13, 60 -80% of cells were differentiated into lipid droplet-containing adipocytes. UCP1 mRNA levels in differentiated cells were similar to the levels found in native subcutaneous adipose tissue. 2  Quantitative Reverse Transcriptase-PCR Analysis-Total RNA was isolated using RNeasy kit (Qiagen, Courtaboeuf, France). Total RNA (1 g) was treated with DNase I (DNase I amplification grade, Invitrogen), then retrotranscribed using random hexamers and Thermoscript reverse transcriptase (Invitrogen). Real time quantitative PCR was performed on GeneAmp 7000 Sequence Detection System using SYBR green chemistry (Applied Biosystems, Courtaboeuf, France) as described (24). Primers were designed using the Primer Express 1.5 software (Table I). Some mRNAs were quantified using Assay-on-Demand gene expression assays (Applied Biosystems). 18 S ribosomal RNA was used as control to normalize gene expression using the ribosomal RNA control Taqman assay kit (Applied Biosystems). Similar results were obtained using SYBR green-and Assay-on-Demand-based detections for the quantification of PGC-1␣ and UCP1 mRNA levels.
Western Blot Analysis-Mitochondria from mouse BAT and human adipocytes were prepared by differential centrifugation in 10 mM Tris, pH 8, 1 mM EDTA, 250 mM sucrose supplemented with a mixture of protease inhibitors (Sigma). Mitochondrial proteins (5 g for human adipocytes and 0.2 g for BAT) were subjected to 10% SDS-PAGE, transferred onto nitrocellulose membrane (Hybond ECL, Amersham Biosciences, Orsay, France), and probed with a polyclonal anti-rat UCP1 antibody (25) and an anti-cytochrome c antibody (Pharmingen, Le Pont de Claix, France). Immunoreactive protein was determined by enhanced chemiluminescence reagent (Amersham Biosciences).
In Vivo Adenovirus Injection in Mouse Fat Pad-Studies with mice followed the INSERM and Louis Bugnard Institute Animal Care Facility guidelines. Male B6D2/JIco mice (24 -30 weeks old, IFFA-CREDO, L'Arbresle, France) were anesthetized with avertin (Sigma). Following dissection of the skin and body wall, one testis with attached epididymal fat pad was pulled out. The adenoviral preparation (1.7 ϫ 10 8 infectious particles) was injected to 6 points in the fat pad. A fat pad was injected with PGC-1␣ adenovirus and the contralateral fat pad with GFP adenovirus. After 5 days, total RNA was prepared from the fat pads for quantitative reverse transcriptase-PCR analyses.
Palmitate Oxidation Experiment-Differentiated human adipocytes were incubated for 3 h in a medium containing Dulbecco's modified Eagle's medium without glucose, 50 mM Hepes, pH 7.8, 1% fatty acidfree bovine serum albumin, 2 mM L-carnitine, 50 M palmitate, and 118 nM [ 14 C]palmitate (850 Ci/mol, Amersham Biosciences). Medium was transferred in a flask with a center well containing Carbosorb E (PerkinElmer Life Sciences, Courtaboeuf, France). 14 CO 2 was liberated by acidification with 5 N HCl and collected overnight on Carbosorb. 14 CO 2 was measured by scintillation counting. The acid-soluble fraction of the medium containing 14 C-labeled ␤-oxidation metabolites were measured by scintillation counting after 1-butanol extraction of palmitate. To inhibit fatty acid oxidation, we used 50 M etomoxir (from Wolfgang Langhans, Swiss Federal Institute of Technology, Zurich, Switzerland). To study the effect of mitochondrial uncoupling, m-chlo-2 C. Tiraby and D. Langin, unpublished data.
Statistical Analysis-Data are expressed as mean Ϯ S.E. Statistical analyses were performed using analysis of variance with least-square difference post-hoc analysis or Student's t test.

The Nuclear Receptor PPAR␥2 and Its Coactivator PGC-1␣
Transactivate the Human UCP1 Promoter-To test the transcriptional coactivation of the human UCP1 promoter by PGC-1␣, we utilized an adenovirus expressing human PGC-1␣ in combination with expression vectors for PPAR␥2 and its partner, RXR␣. PGC-1␣ functions as a transcriptional coactivator for the two nuclear receptors (11,26). The 6300-bp human UCP1 promoter region mediates the stimulation of transcription by TZD in a murine brown adipocyte cell line (21). In simian CV-1 cells, the PPAR␥2/RXR␣ combination had no transactivation potency (Fig. 1). The expression of PGC-1␣ alone induced a modest rise of UCP1 gene transcription. However, the addition of PGC-1␣ to the PPAR␥2/RXR␣ combination led to a marked increase in activity.
PGC-1␣ Induces UCP1 Expression in Human White Adipocyte-Human WAT express low levels of PGC-1␣ (20). We used a human PGC-1␣ adenovirus to increase the expression of the coactivator in primary culture of human subcutaneous adipocytes ( Fig. 2A). Unlike retroviral vector systems that must be used on proliferating preadipocytes, adenoviruses can transduce quiescent mature adipocytes (27). This method permits to avoid the effect of continuous PGC-1␣ expression during adipogenesis. The PGC-1␣ adenovirus efficiently transduced differentiated adipocytes as revealed by GFP labeling of 40 -50% fat cells. Remarkably, adenoviral infection occurred exclusively in differentiated cells (Fig. 2B). There was no GFP staining in undifferentiated fibroblasts. To ascertain that preadipocytes were resistant to adenoviral infection, human fibroblasts at day 3 of culture were infected with PGC-1␣ adenovirus at various multiplicity of infections. No increase in PGC-1␣ mRNA was observed at an m.o.i. of 500. At m.o.i. of 1000 and 2000, PGC-1␣ mRNA were only increased by 3-and 5-fold, respectively, whereas the induction was 100 -150-fold in day 13 differentiated adipocytes at a m.o.i. of 200. The data reveal that human adipocytes possess much more efficient plasma membrane binding and internalization components for human serotype 5 adenovirus than preadipocytes. The effects mediated by PGC-1␣ can thereby be ascribed to its selective overexpression in adipocytes.
The robust overexpression of PGC-1␣ in human adipocytes was accompanied by an induction of UCP1 mRNA expression (Fig. 3A). UCP1 mRNA was barely detectable in cells infected with control GFP adenovirus. The marked increase associated with PGC-1␣ expression was amplified in the presence of PPAR␥ and RXR␣ ligands (p Ͻ 0.01). As shown on Fig. 3B, the mRNA level of UCP1 paralleled that of PGC-1␣ when cells were transduced at different multiplicity of infections. The expression of GFP allowed us to select the transduced cells using flow cytometry (Fig. 3C). Cells infected with PGC-1␣ adenovirus and treated with TZD and RA co-expressed PGC-1␣ and UCP1 mRNA. These data show that the induction of UCP1 is restricted to adipocytes expressing PGC-1␣. Western blot analysis showed an increase of a 32-kDa immunoreactive band corresponding to UCP1 (Fig. 4). Similar results were obtained using an antibody raised against the whole rat protein (25) or an antibody directed against a 19-amino acid C-terminal UCP1 peptide. 2 As for the mRNA level, the addition of PPAR␥/RXR␣ ligands further increased UCP1 protein expression. The amount of UCP1 mRNA and protein in transduced and treated human adipocyte cultures represented 1.0 Ϯ 0.4 and 1.6 Ϯ 0.6% (n ϭ 4) of the corresponding levels in mouse brown fat.
PPAR␥ Cooperates with PGC-1␣ to Induce UCP1 Expression-To determine which PPAR was involved in PGC-1␣ coactivation of UCP1 expression, we tested ligands for the three subtypes of nuclear receptors (Fig. 3D). Rosiglitazone, a PPAR␥ agonist, was a potent stimulator of UCP1 expression. Agonists for PPAR␣ and PPAR␤ and 9-cis-RA had poor inducing potency. The data show the pre-eminent role of the PPAR␥/ PGC-1␣ association for the induction of UCP1 in human white adipocytes.

PGC-1␣-expressing Adipocytes Show Increased Expression of Mitochondrial Proteins and Brown Adipocyte Markers-We
wished to determine whether the up-regulation of UCP1 in TZD-treated PGC-1␣-overexpressing human adipocytes was associated with other metabolically relevant adaptations in gene expression. PGC-1␣ induces expression of components of the mitochondrial respiratory chain in several cell types (11,12). The mRNAs for cytochrome c and cytochrome oxidase 4 were increased (Fig. 5). An induction of cytochrome c was found at the protein level (Fig. 4). We also observed an induction in the mRNA expression of mitofusin 2, a mitochondrial protein essential for mitochondrial network architecture highly expressed in brown adipose tissue (28). Ectopic expression of PGC-1␣ and PPAR␣ in 3T3-L1 murine fibroblasts leads to an increase in mitochondrial fatty acid oxidation enzyme gene expression (29). We therefore determined the mRNA levels for muscle carnitine palmitoyltransferase I, the isoform of CPT1 expressed in human white adipocytes (30), and medium chain acyl-coenzyme A dehydrogenase (MCAD). Treated PGC-1␣-expressing adipocytes showed higher muscle carnitine palmitoyltransferase I and MCAD mRNA levels than control cells. Glycerol kinase activity is very low in WAT while relatively high levels are found in BAT (31). PGC-1␣ in the presence of TZD and RA led to a 7-fold induction of glycerokinase mRNA levels.
PGC-1␣ Does Not Alter TZD Response of Typical White Adipocyte Target Genes-We next investigated the effect of PGC-1␣ on the response of WAT genes regulated by TZD (Fig.  6). Adipocyte lipid-binding protein and cytosolic phosphoenolpyruvate carboxykinase are direct targets of TZD in white adipocytes with identified PPAR␥ responsive elements (32,33). Rosiglitazone increased adipocyte lipid-binding protein and phosphoenolpyruvate carboxykinase 1 mRNA levels in cells transduced with PGC-1␣ or GFP adenovirus. Hence, the expression of PGC-1␣ was not accompanied by a loss of TZD response of target genes.  (34). Compared with the contralateral side injected with GFP adenovirus, the increase in PGC-1␣ mRNA level was accompanied by a more than 10-fold induction of UCP1 mRNA expression (Fig. 3E). The experiment reveals that PGC-1␣ is able to induce UCP1 gene expression in WAT in vivo.

PGC-1␣ Increases Fat Oxidation in Human White
Adipocytes-We hypothesized that an induction of UCP1, proteins of the respiratory chain, and enzymes of fatty acid oxidation may lead to an increased capacity of fat oxidation. Fig. 7 shows that PGC-1␣-expressing human adipocytes had higher total fatty acid oxidation than cells infected with the GFP adenovirus. The effect was mimicked by the chemical mitochondrial uncoupler, m-chlorocarbonylcyanide phenylhydrazone. This data reveals that, in human adipocytes, uncoupling of the respiratory chain is associated with an increase in fatty acid oxidation. There was both an increase in the production of CO 2 and ␤-oxidation metabolites. 2 Addition of rosiglitazone potentiated the effect of PGC-1␣ (p Ͻ 0.01). Compared with basal conditions, total oxidation was doubled. Addition of etomoxir, a CPT1 inhibitor, abolished palmitate oxidation.

DISCUSSION
The transcriptional coactivator PGC-1␣ may play an important role in adaptive thermogenesis in rodents through a positive regulation of mitochondrial proteins (11,35,36). In this study, we show that expression of PGC-1␣ in human white adipocytes specialized in storage of energy induces the expression of UCP1, respiratory chain protein, fatty acid oxidation enzyme, and other brown adipocyte markers. The most salient observation is the marked induction of UCP1 expression. Expression of PGC-1␣ is thus sufficient to trigger the expression of UCP1 in human white adipocytes. If only transduced cells are considered, the level of UCP1 reaches 4 -6% of the UCP1 amount in mouse BAT. Ectopic expression of UCP1 in transgenic mice at 1% of BAT UCP1 content impedes the development of obesity (37). Moreover, PGC-1␣ induces cytochrome c and cytochrome oxidase 4, two respiratory chain proteins. Interestingly, an increase in mitofusin 2 mRNA expression was observed. The protein is highly expressed in mitochondria from brown adipose tissue and skeletal muscle (28). Mitofusin 2 participates in the maintenance of the mitochondrial network architecture and controls mitochondrial metabolism, most notably cellular respiration and mitochondrial proton leak. We also show that TZD and PGC-1␣ stimulate the expression of the mitochondrial fatty acid enzymes CPT1 and MCAD. CPT1 catalyzes the initial reaction in the mitochondrial import of long chain fatty acids, a tightly regulated step in fatty acid utilization. MCAD catalyzes a pivotal reaction of the ␤-oxidation cycle. Other markers of brown adipocytes such as glycerol kinase were induced (31).
The coordinated regulation of gene expression suggests a potential increase in the capacity of fatty acid oxidation. Functional studies corroborate that view. Palmitate oxidation was indeed elevated in the modified adipocytes. These data are the first demonstration of an increase in the fat oxidation capacity of human white fat cells. The increase in CPT1 and MCAD expression is important as the two proteins control limiting steps of ␤-oxidation (38). However, adaptation at the level of the respiratory chain is probably essential. Increase in respiratory chain protein and UCP1 content may lead to an increase of cellular respiration with a stimulation of oxidative phosphorylation and uncoupling. The marked up-regulation of UCP1 and uncoupling capacity is important because it enables the cells to oxidize fatty acids without the kinetic limitations imposed by respiratory control (39). Accordingly, we show that addition of a chemical uncoupler stimulates fatty acid oxidation. In TZD-treated adipocytes expressing PGC-1␣, free fatty acids (e.g. derived from intracellular lipolysis) could furnish NADH and FADH 2 to the respiratory chain through ␤-oxidation and also directly activate UCP1 as occurs during cold exposure in BAT (40). Human white adipocytes can therefore acquire functional features of brown adipocytes. The molecular mechanisms involved in the regulation of the human UCP1 gene has been partially elucidated. A 6300-bp 5Ј-flanking region mediates the positive effects of ␤-adrenergic agonist, RA and TZD (21). The genomic fragment contains a 350-bp enhancer organized as a multipartite response element partially homologous to the mouse and rat enhancers (41). Our data demonstrate that PGC-1␣ can coactivate the PPAR␥2/ RXR␣ heterodimer to stimulate the human UCP1 promoter. Stimulation of transcription is associated with a PPAR␥-dependent increase in UCP1 mRNA and protein levels in human white fat cells. PPAR␣ and PPAR␤ agonists have minor effects. This is different from brown adipocytes where both PPAR␣ and PPAR␥ cooperate with PGC-1␣ to activate UCP1 gene transcription (42,43). The regulation of classical white adipocyte PPAR␥-responsive genes is not altered by PGC-1␣ because adipocyte lipid-binding protein and phosphoenolpyruvate carboxykinase induction by TZD is preserved. Promoter-specific interactions between PPAR␥ and coactivators may partially underlie the similarities and differences in gene expression between brown and white adipocytes.
In response to cold, appearance of brown fat cells is observed in mouse visceral WAT (44). An unanswered question is the origin of the novel brown fat cells. The cells could originate from differentiation of a specific pool of precursor cells already present in WAT. Accordingly, the presence of UCP1 was detected in about 10% of adipocytes differentiated from a Siberian dwarf hamster white adipocyte precursor pool (45). It has also been proposed that some unilocular white adipocytes are "masked" brown fat cells that revert to the brown adipocyte phenotype (46). Finally, brown fat cells could derive from direct conversion of white adipocytes (47,48). As PGC-1␣ was expressed exclusively in fully differentiated white fat cells and not throughout the differentiation process, our data reveal that transdifferentiation of mature white adipocytes into UCP1expressing cells can be activated in human WAT. The results could have implications on drug discovery strategy as it brings the experimental proof for conversion of mature white adipocytes into brown-like fat cells.
The notion of an activation of nonshivering thermogenesis in BAT to prevent excessive fat storage in situations of high calorie intake was first introduced by Rothwell and Stock (49). An apparent paradox comes from UCP1-deficient mice that show cold intolerance but do not become obese (50,51). However, emergence of brown fat cells in white fat depots is asso- FIG. 4. Western blot analysis of UCP1 and cytochrome c expression. Mitochondrial proteins were prepared from mouse BAT (0.2 g) and human white adipocytes (5 g). Adipocytes were treated with rosiglitazone (Rosi) and 9-cis-retinoic acid (RA). ciated with a lean phenotype in several transgenic models (52)(53)(54)(55). The mice have enhanced metabolic rate and insulin sensitivity and, are protected against diet-induced obesity. Furthermore, transgenic mice expressing UCP1 in WAT are protected against genetic and dietary obesity and show an increase in WAT oxygen consumption (37,56). BAT of these mice is atrophied and the animals are cold-sensitive (57). UCP1 levels are low in retroperitoneal fat pads of genetically obese rats and mice (34,46). It is therefore possible that the role of UCP1 and brown fat cells differ according to the location in BAT or WAT, the latter being associated with obesity resistance.
The cooperation between PPAR␥ and PGC-1␣ may explain the effects of TZD on thermogenesis. In vivo treatments with TZD have been reported to increase BAT mass (58,59). The increased formation of BAT is accompanied by an increase in UCP1 expression (59 -61). The effect is observed in lean animals but also in genetically obese mice that are characterized by a defect in adaptive thermogenesis. A TZD, NC-2100, has been shown to promote a robust antidiabetic effect on KKAy obese mice without the increase in the weight of white fat depots reported with classical TZD (62). The lack of weight gain may partially be explained by an induction of UCP1 in WAT that is stronger for this compound than for classical TZD. Ectopic expression of PGC-1␣ in white adipocytes is therefore sufficient to direct PPAR␥ toward a transcriptional program linked to energy dissipation besides its classical role in adipogenesis and maintenance of the white adipocyte phenotype (63,64). The in vivo data in rodents and our results suggest that a combination of TZD and inducers of PGC-1␣ expression may enhance the oxidative capacity of human WAT.
Obesity is commonly seen as a disorder of energy balance, where energy intake exceeds energy expenditure. Mobilization of WAT without use of released fatty acids may be deleterious, both as the excess calories will be deposited in other organs as seen in lipoatrophic models (65,66) and because excess fatty acids may play a role in the development of insulin resistance and cardiovascular complications (67). Therapeutic strategies to increase the expression and activity of PGC-1␣ in WAT could contribute to the induction of UCP1 expression and fatty acid oxidation leading to a decrease in fat mass.