Browning of White Adipose Cells by Intermediate Metabolites: An Adaptive Mechanism to Alleviate Redox Pressure

Recent discoveries in adult humans have renewed the vision of brown adipose tissue (BAT) as a therapeutic target for treating metabolic disorders such as type 2 diabetes and obesity (1). Whereas white adipose tissue (WAT) represents the main energy store, BAT dissipates energy as heat by uncoupling the mitochondrial respiratory chain from ATP formation through expression of uncoupling protein (UCP)1 (2). Thus, BAT plays a key role in diet and nonshivering thermogenesis (3). While it was assumed for many years that in humans, BAT was only present in the neonate (with some exceptions such as cold-exposed workers or patients suffering from phaeochromocytoma [4]), recent imaging studies revealed the presence of substantial deposits of UCP1-expressing adipocytes in human healthy adults, whose activity is negatively correlated with BMI (5).

Certain WAT depots are readily able to convert to a “brown-like” state with prolonged cold exposure or treatment with β-adrenergic compounds (6,7). Several hormones, enzymes, transcription factors, and microRNAs have recently been shown to drive the emergence of inducible brown adipocytes in WAT (8). As increase clearance and use of nutrients by brown adipocytes reduce the excess of triglycerides and glucose and confer beneficial metabolic effects and/or protection from obesity in mice (9–11), increasing the activity and/or emergence of UCP1-positive cells holds promise for the treatment of metabolic diseases.

The influence of the metabolic environment on the browning process is to date poorly understood. Both white and brown adipose cells express metabolite receptors and transporters, which may act as sensors of their metabolic environment. Proton-linked monocarboxylate transporters (MCTs) (12), which drive lactate, pyruvate, and ketone body transport across plasma membranes, are present on the surface of rodent (13–15) and human (16) adipocytes. Notably, the MCT1 isoform, mainly involved in lactate import into cells, strongly increases in BAT after exercise (15). WAT adipocytes also express the G-protein–coupled receptors (GPRs) GPR81 and GPR109A, which are activated by lactate and the ketone body β-hydroxybutyrate (βHB), respectively (17).

Among intermediate metabolites, lactate mediates a large intercellular and interorgan metabolic interplay (18). Lactate uptake, which induces changes in cellular redox state, is associated with neoglucogenesis or ATP production (19). Whereas lactate has been considered for a long time as a glycolytic waste product, it acts as a metabolic fuel supporting organ (Cori cycle) and cell interactions between astrocytes and neurons (20), as well as between glycolytic and oxidative cancer cells (21). Furthermore, its role as a signaling molecule (termed “lactormone” [19]) has recently been shown (22–26).

Given recent evidence demonstrating that expression of MCT in adipose tissues is controlled by physiological stimuli of browning (15), we aimed to investigate whether lactate could directly regulate browning remodeling of white adipocytes. Here, we show that lactate strongly increases thermogenic gene expression in mouse and human white adipose cells, this effect being dependent on the presence of active peroxisome proliferator–activated receptor (PPAR)γ signaling. Importantly, we demonstrate that the lactate effect on Ucp1 is mediated by intracellular redox modifications as a result of lactate transport through MCT. We also show that βHB, another metabolite having the same consequence on cellular redox state, is also a strong browning enhancer.

All studies were carried out using male C57BL/6J mice obtained from the Harlan Laboratory. Pparα-null mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Animals were housed in a controlled environment (12-h light/dark cycles at 21°C) with unrestricted access to water and a standard chow diet (Usine d'Alimentation Rationnelle, Villemoisson sur Orge, France) in a pathogen-free animal facility (IFR150) and killed by cervical dislocation. All experimental procedures were done in compliance with French Ministry of Agriculture regulations for animal experimentation.

White stromal vascular fractions from 6- to 8-week-old mice were obtained as previously described (27). After centrifugation, stromal vascular fraction cells resuspended in culture medium (αMEM plus 0.25 units/mL amphotericin, 100 units/mL penicillin, 100 mg/mL streptomycin, biotin, ascorbic acid, panthotenic acid, and 10% newborn calf serum) were plated and rinsed with PBS 3 h after plating. Remaining adherent adipose-derived stem/stromal cells were grown to confluence and exposed to the adipogenic cocktail containing 5 µg/mL insulin, 2 ng/mL T3, 33.3 nmol/L dexamethasone, 10 µg/mL transferrine, and 1 μmol/L rosiglitazone in complete medium. Adipocytes differentiated for 7 days were treated with different compounds at the time and concentrations as indicated in the figure legends.

Human adipocyte progenitors were derived from induced pluripotent stem cells and cultivated as already described [28,29,29(a)]. For adipocyte differentiation, cells were cultured in proliferation medium, which was supplemented with 0.5 mmol/L isobutylmethylxanthine, 0.25 μmol/L dexamethasone, 0.2 nmol/L T3, 1 μg/mL insulin, and 1 μmol/L BRL49653 once cells reached confluence.

We extracted proteins and performed Western blotting as previously described (30). Sources of antibodies are UCP1 (UCP11-A, diluted 1:2,000; Alpha Diagnostic), MCT4 (sc-50329, diluted 1:500; Santa Cruz Biotechnology) or actin (A5541, diluted 1:5,000; Sigma-Aldrich).

Six-week-old mice were randomly divided into four groups and injected daily (200 μL i.p.) with either 5% DMSO/lutrol F127 2% (BASF France, Levallois-Perret, France) and PBS 93% (v/v/v), rosiglitazone (250 μg/mice), l-sodium lactate (375 μmol/mice), and a combination of both molecules. After 11 days of injection and overnight fasting, mice were killed by cervical dislocation and inguinal and interscapular fat pads were harvested. Concerning UCP1 immunohistochemistry, mice were anesthetized by an injection of ketamine (100 mg/kg i.p.)-xylazine (10 mg/kg) before receiving an intracardiac perfusion of 3.6% formaldehyde. Total inguinal depots were removed, postfixed at 4°C overnight, and rinsed in PBS.

Sections of inguinal depots (200 μm) or differentiated adipocytes cultivated in Lab-Teks were blocked and permeabilized for 30 min at room temperature in PBS–2% horse serum–0.2% triton before incubation with sheep anti-UCP1 primary antibody (generous gift from Daniel Ricquier, Institut Cochin, Paris, France; Institut National de la Santé et de la Recherche Médicale, Paris, France; Centre National de la Recherche Scientifique, Paris, France; and Université Paris Descartes, Paris, France) diluted at 1:5,000. After washing steps, samples were incubated with Alexa Fluor 488 secondary antibody (Invitrogen, Cergy Pontoise, France). Adipocytes were visualized by staining lipid droplets with 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene-3-dodecanoic acid (BODIPY) 558/568 C12 diluted at 1:2,000. Nuclei were stained with DAPI diluted at 1:10,000. Fluorescence analysis was performed using a multiphoton laser scanning microscope (Zeiss LSM 510 NLO confocal microscope equipped with a femtosecond Coherent pulsed laser).

Plasma and extracellular lactate levels were measured with the Lactate Pro test meter (Arkray). Glucose concentrations in cell supernatants were measured using Contour TS (Bayer). Values were normalized by cellular protein content of the feeder cell layer (differentiated adipocytes).

The whole-cell layer of differentiated adipocytes was harvested and incubated in culture medium without serum and supplemented with 4% free fatty acid BSA in a magnetically stirred oxygen electrode chamber set to 37°C. Oxygen consumption was measured polarographically using a Clark oxygen electrode (Oxygraph-2k Oroboros). The chamber was closed, and the cells were incubated to determine the basal respiratory rate. Oligomycin (2 µg/mL) was then added to measure the coupled respiration rate and antimycin (20 µmol/L) was added to measure mitochondrial-derived oxygen consumption. Oxygen consumption rate was determined from the slope of a plot of O₂ concentration versus time.

NADH-to-NAD⁺ ratio was assessed by the EnzyChrom NAD⁺/NADH Assay kit (BioAssay Systems).

Samples of fresh supernatant of differentiated adipocytes were harvested, and glycerol was measured using the Triglycérides Enzymatiques PAP 150 (bioMérieux).

At 80% confluence, murine adipose-derived stem/stromal cells were dissociated in 0.25% trypsin EDTA and numbered. For each nucleofection assay, 2 × 10⁶ cells were nucleofected (Amaxa Cell line Nucleofector Kit V; Amaxa) with 1.6 μmol/L of scramble siRNA (AllStars Negative Control siRNA; Qiagen), Mct1 siRNA (Mm_Slc16a1_2; Qiagen), or Mct4 siRNA (Mm_Slc16a3_5; Qiagen) with the Nucleofector II device (X005 protocol). After 4 days of differentiation, cells were treated with 25 mmol/L lactate for 24 h.

Total RNA from culture cells was isolated using an Qiagen RNeasy kit (Qiagen). For mouse tissues, total RNA was isolated by Qiazol extraction and purification was done using RNeasy minicolumns. For quantitative real-time PCR analysis, 300–1,000 ng total RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription kit (Life Technologies/Applied Biosystem), SYBR Green PCR Master Mix (Life Technologies/Applied Biosystem), and 300 nmol/L primers on an Applied Biosystem StepOne instrument. Relative gene expression was calculated by the ∆∆CT method and normalized to 36B4 for mouse and TBP for human analysis. Primers are listed in Table 1.

All results are expressed as mean ± SEM from at least three individual experiments. Unpaired t test was used to calculate final P values.

As expected, higher expression of Ucp1 was detected in BAT compared with white adipose depots (Fig. 1A). We compared the expression of Mct1 and Mct4, the main MCTs enabling lactate uptake and export, respectively (21,22,31), in both brown and white adipose depots. Higher amount of Mct1 mRNA was found in BAT compared with WAT depots (Fig. 1B). Conversely, Mct4 mRNA levels were greater in WAT, especially in the low Ucp1-expressing epididymal fat pad, compared with BAT (Fig. 1C). Furthermore, in mice exposed to 4°C for 1–7 days, we observed, in addition to the increase in Ucp1 expression in both BAT and inguinal depots (Fig. 1D), an increase in the expression of the lactate-importing isoform Mct1 (Fig. 1E) without any change in Mct4 expression (Fig. 1F). In addition, we found that plasma lactate levels significantly rose after 24 h and 72 h of cold exposure and returned to basal levels after 7 days of cold exposure (Fig. 1G). These data show that brown and white adipose depots display different profiles for MCT expression and that in vivo browning induced by cold acclimation is associated with an increased lactate import system in adipose depots together with transient modification of plasma lactate levels.

We then directly assessed the effects of lactate on adipocyte biology. Acute 48-h treatment of differentiated adipocytes isolated from inguinal fat pad with lactate resulted in a robust increase of Ucp1 mRNA levels together with an upregulation of the expression of additional key thermogenic genes such as Cidea, Fgf21, and Hoxc9 without any significant effect on other genes involved in brown fat cells function such as Ppargc1a, Prdm16, Tbx1, Zic1, and Lhx8 (Fig. 2A). This effect is specific to Ucp1, as no change was observed for Ucp2 (Fig. 2A). Lactate treatment also increased the expression of the mitochondrial marker Cox7a1 as well as the fatty acid oxidation marker Cpt1b (Fig. 2A). While no effect was observed for Fabp4 or Lpl adipogenic genes, lactate significantly increased the expression of the nuclear receptor Pparg, a master regulator of both white and brown adipogenesis (Fig. 2A). The increase of Ucp1 mRNA levels by lactate was associated with an increase in UCP1 content as shown by Western blot (Fig. 2B) (in a dose-dependent manner, with an effect starting at 10 mmol/L [data not shown]) and by immunofluorescence (Fig. 2C), which showed that lactate increased both the number of UCP1-positive cells and the intensity of UCP1 staining per cell. While no difference was observed concerning basal respiratory rate (data not shown), we found a higher uncoupled respiration in lactate-treated cells (respiration insensitive to oligomycin; 78 ± 9.1% in lactate-treated cells compared with 27.2 ± 11% in control cells [P < 0.05]), demonstrating the functionality of UCP1. Further highlighting the brown-like phenotype of lactate-treated adipocytes, we found higher glucose consumption rate in lactate-treated cells compared with control cells (1.12 ± 0.04-fold increase compared with control cells [P < 0.05]).

Lactate-increased Ucp1 expression was already observed after 24 h of treatment, in a dose-dependent manner (Fig. 2D), with an effect starting at 5 mmol/L and reaching its maximum at 25 mmol/L. This lactate-induced Ucp1 gene expression was not due to osmotic perturbations, as mannitol did not upregulate Ucp1 (data not shown). Importantly, lactate-mediated increase in thermogenic gene expression also occurred in human-induced pluripotent stem-derived adipocytes (29) (Fig. 2E), demonstrating lactate effect on human cells. Together, these data reveal that lactate is a strong inducer of brown adipose gene expression in white adipose cells, and this effect is conserved between species.

During our investigation of lactate-regulated gene expression in white adipose cells, we observed an increase in the expression of Glut1, Vegf, and Cox4i2 (Fig. 3A). These genes are all targets of the hypoxia-inducible transcription factor hypoxia-inducible factor (HIF)-1α (32,33), which has been shown to be stabilized by lactate (22). Our observation that DMOG, a prolyl-4-hydroxylase inhibitor that stabilizes HIF-1α (32), increased Glut1, Vegf, and Cox4i2 expression as expected but not Ucp1 expression (and, rather, downregulated it [Fig. 3B]), together with the fact that incubation of white adipocytes under 1% O₂ for 24 h did not upregulate Ucp1 expression (data not shown), enabled us to exclude any role for HIF-1α in the control of Ucp1 expression by lactate. As Ucp1 expression is regulated by PPARα and PPARγ signaling pathways, we tested their involvement in lactate effect. The increase in Ucp1 expression by lactate similarly occurred in Ppara knockout versus wild-type adipocytes (Fig. 3C), ruling out a role of PPARα. In contrast, Ucp1 expression in white adipose cells required the presence of the PPARγ ligand rosiglitazone (as already demonstrated [34]), in both basal and lactate-treatment conditions (Fig. 3D) and is decreased by the PPARγ antagonist GW9662 (Fig. 3E). As the expression of Fabp4 and Adipoq, two well-described PPARγ targets, was not upregulated by lactate treatment (Supplementary Fig. 1), we concluded that lactate requires a functional PPARγ signaling to regulate Ucp1 expression but does not activate PPARγ directly.

To determine whether lactate-induced browning also occurs in vivo, we treated mice with daily intraperitoneal injection of lactate for 11 consecutive days. As shown in Fig. 4A and B, lactate treatment by itself did not change Ucp1 expression in inguinal or in interscapular brown adipose tissue (iBAT) fat pads. We then injected lactate in rosiglitazone-treated mice, as this antidiabetes compound had previously been shown to favor the browning process. As previously described (35), rosiglitazone-treated mice displayed an increase in Ucp1 expression in inguinal (Fig. 4A), without any significant difference in Ucp1 expression in iBAT (Fig. 4B). Strikingly, lactate coinjection with rosiglitazone further strongly enhanced both Ucp1 and Cidea expression compared with rosiglitazone-injected mice (Fig. 4A) as well as mRNA levels of Hoxc9, Cox7a1, Cpt1b, Acox, and Pdk4 (Supplementary Fig. 2). This browning effect is specific to subcutaneous white adipose depot, as it did not occur in iBAT (Fig. 4B). The strong browning effect of lactate and rosiglitazone cotreatment on Ucp1 expression was associated with a high increase in UCP1 content (Fig. 4C). High magnification shows that UCP1-positive cells are multilocular adipocytes, whereas unilocular adipocytes are UCP1 negative (Fig. 4D). Together, these data demonstrate that lactate induces the emergence of inducible brown adipocytes in white fat depots of mice treated with a PPARγ agonist.

To gain new insight into the mechanisms underlying lactate-induced Ucp1 expression, we first investigated the role of the lactate receptor GPR81. Increasing doses of the GPR81 agonist 3,5-Dihydroxybenzoic acid did not affect Ucp1 expression (Supplementary Fig. 3A), although this compound was effective in decreasing isoproterenol-induced lipolysis (Supplementary Fig. 3B), as already described (36). To address the putative involvement of MCT in lactate-mediated increase in Ucp1 expression, we used two distinct MCT pharmacological inhibitors that have been shown to reduce lactate influx in several cell types (21,22,37). We found that both α-cyano-4-hydroxycinnamate and phloretin totally abrogated lactate-induced Ucp1 expression (Fig. 5A). Interestingly, these inhibitors also decreased Ucp1 expression in basal conditions (without extracellular lactate addition), suggesting that lactate endogenously produced and consumed by cells controls Ucp1 expression. We then targeted Mct1 and Mct4 expression using an RNA interference approach. While downregulating Mct1 expression induced no effect on Ucp1 mRNA levels (data not shown), Ucp1 expression was very sensitive to Mct4 knockdown (consequences of Mct4 knockdown are shown at mRNA levels [Fig. 5B], protein levels [Fig. 5C], and functional levels assessed by lactate measurement in cell supernatants [Fig. 5D]). Indeed, reducing expression of the lactate-exporting MCT4 isoform strongly enhances Ucp1 expression in both control and lactate-treated conditions (Fig. 5E), suggesting that intracellular lactate levels dictate Ucp1 expression in white adipocytes. Altogether, these data show that lactate regulates Ucp1 expression in adipocytes via MCT-mediated transport and in a GPR81-independent manner.

Among the diverse consequences of increasing intracellular lactate levels through MCT, redox state (i.e., NADH-to-NAD⁺ ratio) might be changed through lactate conversion into pyruvate by the isoform B of the lactate dehydrogenase (Fig. 6A). We then tested the effect of the ketone body βHB, another monocarboxylate able to similarly affect redox state (Fig. 6A), on Ucp1 expression. Strikingly, treatment of differentiated white adipocytes with βHB strongly increased expression of both Ucp1 and Cidea in a dose-dependent manner (Fig. 6B). In contrast, neither acetoacetate nor pyruvate displayed similar effects (Fig. 6C and D). The similar effect of lactate and βHB and the lack of effect of acetoacetate and pyruvate plus the increased NADH-to-NAD⁺ ratio in lactate- or βHB-treated adipocytes (Fig. 6E) clearly suggested a role for cellular redox state in the control of Ucp1 expression. We then cotreated cells with lactate and increasing concentrations of pyruvate to modulate intracellular redox state according to the law of mass action (Fig. 6A). Clearly, addition of pyruvate significantly decreased lactate-induced Ucp1 expression (Fig. 6F), as did the addition of acetoacetate in βHB-treated cells (Fig. 6G). Furthermore, addition of the uncoupling agent CCCP that decreases redox state by dissipating mitochondrial membrane potential abrogated both lactate- and βHB-induced Ucp1 upregulation (Fig. 6H). Altogether, these data demonstrate that cellular redox change induced by lactate and βHB triggers their effect on Ucp1 expression.

Although no metabolite has previously been shown to regulate the browning process, we demonstrate here that lactate drives the fate of white adipose murine and human cells toward a brown and oxidative phenotype through sharp and rapid changes in the expression of several brown genes including Ucp1. We show that this lactate-based browning remodeling is mediated by intracellular redox modifications and that metabolites that similarly impact redox state such as the ketone body βHB also constitute strong browning inducers. We propose that under these conditions, induction of UCP1 constitutes an adaptive mechanism to alleviate redox pressure in adipocytes (Fig. 7).

Whereas lactate has been considered for a long time as a glycolytic waste product, its role as a true signaling molecule (termed “lactormone” [19]) has recently been emphasized. At physiological concentrations (from 10 mmol/L in plasma after acute exercise to 25 mmol/L [38] or even 40 mmol/L in muscle or certain tumors [39], respectively), lactate has been shown to specifically control gene expression in several cell types (22–26). Notably, lactate controls expression of several proteins involved in mitochondrial activity and biogenesis (24). Here, we show that lactate is also a strong enhancer of thermogenic gene expression. This effect starts at 5 mmol/L and is highly specific because it did not affect Ucp2. Lactate can act through MCT transporters (12) and membrane receptors (17). In particular, lactate mediates the antilipolytic action of insulin through activation of the G_i-coupled receptor GPR81 by inducing a decrease in cAMP levels (40). We show here that although activation of GPR81 inhibits isoproterenol-induced lipolysis as already described (36), it does not induce Ucp1 expression. In any event, it is hardly conceivable that activation of G_i-coupled GPR81, which decreases cAMP levels, contributes to Ucp1 expression given the cAMP-dependent regulation of Ucp1 expression under noradrenergic stimulation. Conversely, we show by molecular and knockdown approaches that lactate transport through MCT is critical for the regulation of Ucp1 expression. Whereas the two pharmacological inhibitors used, described as inhibitors of lactate import into several cell types (21,22,37), very efficiently reduce lactate-mediated Ucp1 upregulation, downregulating Mct1 displayed few effects on Ucp1 levels; this, however, may be due to insufficient knockdown of the MCT1 protein and/or its high affinity for lactate (Km ∼3.5 mmol/L). In contrast, Ucp1 expression is very sensitive to Mct4 knockdown, as we found that reducing expression of the lactate exporting MCT4 isoform (which has a low affinity for lactate [Km ∼28 mmol/L]) strongly enhances Ucp1 expression. The basal expression of Ucp1 (without lactate addition) is very sensitive to both pharmacological inhibitors and Mct4 knockdown, revealing an autocrine/paracrine mechanism by which lactate produced and consumed by cells tightly controls Ucp1 expression. All these in vitro data are consistent with our in vivo findings demonstrating that 1) BAT displays higher Mct1-to-Mct4 ratio compared with low Ucp1-expressing white adipose depots, 2) browning induced by cold acclimation is associated with an increase in the lactate-importing Mct1 isoform that parallels Ucp1 upregulation, and 3) plasma lactate levels showed a transient increase under cold-induced browning conditions.

We found that lactate increases expression of several HIF-1α target genes, these findings being in total agreement with recent data demonstrating that lactate stabilizes HIF-1α through its MCT-mediated intracellular transport and direct inhibition of prolyl hydroxylase activity (22). However, we show that HIF-1α signaling is not involved in Ucp1 expression. We also show that lactate regulates Ucp1 expression in a PPARα-independent manner but required an active PPARγ signaling pathway. In addition to these in vitro findings, we also show that the combination of lactate and the antidiabetes compound rosiglitazone constitutes a strong inducer of brown-like adipogenesis (and increased expression of some mitochondrial and β-oxidation markers), in white adipose depot, demonstrating that lactate controls the browning process in vivo.

After its transport inside the cell, lactate is converted into pyruvate by the lactate dehydrogenase following the reaction: lactate + NAD⁺ <=> pyruvate + NADH. Manipulation of the lactate-to-pyruvate ratio directly changes the cytoplasmic NADH-to-NAD⁺ ratio by the law of mass action. Spectacularly, the browning effect of lactate can be blunted with pyruvate; this undoubtedly demonstrates the key role of NADH-to-NAD⁺ ratio on Ucp1 expression. Similar conclusions on the impact of NADH-to-NAD⁺ ratio can be drawn from the comparison between the effects of acetoacetate and βHB. Furthermore, it is remarkable that the uncoupling of mitochondria respiration by a chemical uncoupler mimicking UCP1 activity and thus consuming NADH alleviates the lactate-mediated increase in Ucp1 expression. One can postulate that cells having high UCP1 activity and thus having a high capacity to dissipate redox pressure will become less responsive to lactate and βHB treatments than cells having low UCP1 expression and activity. Several protein sensors whose activities are dependent on the NADH-to-NAD⁺ ratio synchronize cell signaling and transcriptional events with the cellular metabolic state (41). Among them, some regulate the expression of white and brown fat–selective gene programs such as the transcriptional coregulator COOH-terminal–binding protein-1 (CtBP-1) (42,43), as well as the NAD⁺-dependent Sirtuin-1 enzyme (44). Among the complex and redundant redox-sensitive signaling pathways, the identification of the lactate-redox–sensitive target(s) requires further investigations.

Our findings suggest that the browning remodeling of white adipose tissue, described in situations where such intermediate metabolites are released, could be due, at least partly, to their direct effects on adipose cells. Our in vitro imaging suggests that lactate-induced browning probably corresponds to both the emergence of new UCP1 positive cells and activation of preexisting low-UCP1–expressing cells. During exercise, beside the role of irisin (45), any increase in lactate could contribute to the browning remodeling. This is also in agreement with a recent demonstration that a ketone ester–enriched diet activates BAT and increases Ucp1 expression in white adipose depot of C57BL/6J mice (46). It is also noteworthy that βHB constitutes a major energy metabolite for the newborn at a time where brown fat thermogenesis is crucial (47). Our findings bring very new insights in the browning field, as they demonstrate for the first time that specific metabolites have a drastic and direct impact on this phenomenon. Furthermore, because lactate or βHB promotes an oxidative phenotype with uncoupled mitochondria and elevated oxidation rate that in turn will decrease the redox pressure, browning remodeling of white adipose tissue and redox-dependent increase in Ucp1 expression appears as an adaptive mechanism to alleviate redox pressure. This corresponds to an unexpected and new function for the development of these brown-like adipocytes and opens new perspectives for the control of adipose tissue browning and its physiological relevance.

Acknowledgments. This work is dedicated to Xavier Leverve (Université de Grenoble) and, for our challenging and friendly discussions, to Michel Rigoulet (Institut de Biochimie et Génétique Cellulaires, UMR 5095, Université Bordeaux Segalen). We are grateful to Institut des Technologies Avancées en Sciences du Vivant (Unité Mixte de Service 3039) and to Institut des Maladies Métaboliques et Cardiovasculaires/UMR 1048, Plateforme Génome et Transcriptome du Génopole Toulouse. We especially thank Justine Caturla; Christophe Guissard, Clément Vecchi, and Pascale Guillou (UMR 5273 CNRS UPS EFS INSERM U1031); and Célia Bettiol (former student of UMR 5273 CNRS UPS EFS INSERM U1031) for excellent technical assistance.

Funding. This work was supported by the European Union FP7 project DIABAT (HEALTH-F2-2011-278373), the European Union FP7 project METABOSTEM (PCIG9-GA-2011-293720), and the ANR Safe.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.C. designed the experiments, supervised experiments, researched and analyzed data, and wrote and edited the manuscript. Y.J. and S.B.-M. designed the experiments, researched and analyzed data, and reviewed the manuscript. M.A., V.C., E.A., C.B., R.W., A.G., B.W., P.V., and K.L. researched and discussed data. P.C., C.M., C.D., and F.V. contributed to the experiments, discussed data, and reviewed the manuscript. L.C. designed the experiments, analyzed data, and wrote and reviewed the manuscript. A.C., Y.J., and L.C. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.