Am J Physiol Endocrinol Metab 309: E105–E114, 2015. First published May 26, 2015; doi:10.1152/ajpendo.00518.2014.

CALL FOR PAPERS

Novel Aspects of Adipocyte Biology

The nuclear retinoid-related orphan receptor-␣ regulates adipose tissue glyceroneogenesis in addition to hepatic gluconeogenesis Sarah Kadiri,1,2 Chloé Monnier,1,2 Munkhzul Ganbold,1,2 Tatiana Ledent,1,2 Jacqueline Capeau,1,2 and Bénédicte Antoine1,2,3 1

Institut National de la Santé et de la Recherche Médicale, U938, Paris, France; 2Sorbonne Université, University Pierre et Marie Curie; Univ Paris-6, UMR_S 938, l’Institut de Cardiométabolisme et Nutrition, Paris, France; and 3Centre National de la Recherche Scientifique, UMR_S 938, Paris, France

Submitted 11 November 2014; accepted in final form 17 May 2015

Kadiri S, Monnier C, Ganbold M, Ledent T, Capeau J, Antoine B. The nuclear retinoid-related orphan receptor-␣ regulates adipose tissue glyceroneogenesis in addition to hepatic gluconeogenesis. Am J Physiol Endocrinol Metab 309: E105–E114, 2015. First published May 26, 2015; doi:10.1152/ajpendo.00518.2014.—Circadian rhythms have an essential role in feeding behavior and metabolism. ROR␣ is a nuclear receptor involved in the interface of the circadian system and metabolism. The adipocyte glyceroneogenesis pathway derives free fatty acids (FFA) liberated by lipolysis to reesterification into triglycerides, thus regulating FFA homeostasis and fat mass. Glyceroneogenesis shares with hepatic gluconeogenesis the key enzyme phosphoenolpyruvate carboxykinase c (PEPCKc), whose gene is a ROR␣ target in the liver. ROR␣-deficient mice (staggerer, RORsg/sg) have been shown to exhibit a lean phenotype and fasting hypoglycemia for unsolved reasons. In the present study, we investigated whether adipocyte glyceroneogenesis might also be a target pathway of ROR␣, and we further evaluated the role of ROR␣ in hepatocyte gluconeogenesis. In vivo investigations comparing RORsg/sg mice with their wild-type (WT) littermates under fasting conditions demonstrated that, in the absence of ROR␣, the release of FFA into the bloodstream was altered and the rise in glycemia in response to pyruvate reduced. The functional analysis of each pathway, performed in adipose tissue or liver explants, confirmed the impairment of adipocyte glyceroneogenesis and liver gluconeogenesis in the RORsg/sg mice; these reductions of FFA reesterification or glucose production were associated with decreases in PEPCKc mRNA and protein levels. Treatment of explants with ROR␣ agonist or antagonist enhanced or inhibited these pathways, respectively, in tissues isolated from WT but not RORsg/sg mice. Our results indicated that both adipocyte glyceroneogenesis and hepatocyte gluconeogenesis were regulated by ROR␣. This study demonstrates the physiological function of ROR␣ in regulating both glucose and FFA homeostasis. glyceroneogenesis; phosphoenolpyruvate carboxykinase c; nuclear retinoid-related orphan receptor-␣; adipose tissue RETINOID-RELATED ORPHAN RECEPTORS (RORs) are members of the nuclear receptor family. The ROR family comprises three members: ROR␣ (NR1F1), ROR␤ (NR1F2), and ROR␥ (NR1F3), with ROR␣ being the most abundant isoform present in the adipose tissue (18). RORs regulate gene transcription by binding to specific DNA response elements (RORE) consisting of the consensus RGGTCA core motif. Reciprocally, the nu-

Address for reprint requests and other correspondence: B. Antoine, Faculté de Médecine site Saint-Antoine, 27 Rue Chaligny, 75012 Paris, France (e-mail: [email protected]). http://www.ajpendo.org

clear receptor Rev-erb␣ (NR1D1) acts as a transcriptional repressor by competing with RORs for RORE binding and thereby antagonizes ROR action. Indeed, both receptors are often coexpressed in the same tissues (14). ROR␣ and Rev-erb␣ are involved in many pathways implicated in various physiological functions, including immune function, circadian rhythms, and metabolism. They appear to be pivotal players at the interface between the circadian system and metabolism (49). Among the direct target genes of ROR␣ and Rev-erb␣ are several clock genes such as Bmal1, thus ensuring a fine tuning of circadian rhythms, as well as some metabolic genes involved in the control of cholesterol and bile acid metabolism (11, 29) apolipoprotein synthesis (39), lipogenesis in the liver (26), energy expenditure in the muscle (25), and glucose metabolism (see Ref. 37 for review). Hepatic glucose production was first shown to be regulated by Rev-erb␣, which represses the expression of the G6PC and PCK1 genes encoding the two major enzymes involved in gluconeogenesis [glucose-6-phosphatase and the cytosolic form of phosphoenolpyruvate carboxykinase (PEPCKc)] in HepG2 cells (50). Then, a synthetic ROR␣-inverse agonist was reported to inhibit the expression of these two gluconeogenic genes and, in vivo, when administrated to obese mice, to reduce plasma glucose levels following a pyruvate tolerance test (PTT) (24). These observations suggested a role for ROR␣ as an activator of gluconeogenesis, whereas Rev-erb␣ would be an inhibitor. The phenotypes of the knockout mice fit with the reported opposed roles of ROR␣ and Rev-erb␣ on metabolism; mice that are deficient in Rev-erb␣ have increased plasma glucose levels (9), whereas ROR␣-deficient staggerer (RORsg/sg) mice are hypoglycemic during fasting (27). In adipose tissue, there is a truncated pathway of gluconeogenesis, glyceroneogenesis, that controls free fatty acid (FFA) homeostasis during fasting periods. Studies with a broad variety of mammalian species (including human) have demonstrated that a sizeable fraction of FFA released from adipose tissue during fasting is reesterified into triglycerides (TG) in adipocytes (40). This cycle provides a way for the organism to adapt for the exact energy needs of an individual. The reesterification of FFA within adipocytes requires the generation of glycerol 3-phosphate (G3P), and this occurs through glyceroneogenesis, which is the de novo synthesis of G3P from precursors such as lactate and pyruvate. In vivo, glyceroneogenesis was shown to be the predominant pathway for TG synthesis in rodents (34) and an important

0193-1849/15 Copyright © 2015 the American Physiological Society

E105

E106

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

regulator of body fat mass. Indeed, the adipose tissue-specific invalidation of its key gene PCK1 generated mice that have a reduced fat mass (36). We provided evidence that the hypolipidemic action of glitazones was due in part to their ability to induce transcription of the PCK1 gene in rodent and human adipose tissue, thus increasing the rate of glyceroneogenesis and decreasing the plasmatic FFA level (3, 30, 44). Very recently, a type 2 diabetes risk locus was identified in the human peroxisome proliferator-activated receptor (PPAR)␥2 gene promoter, which decreased both PCK1 expression and pyruvate incorporation in adipose tissue and increased FFA release (8), confirming the importance of glyceroneogenesis in the physiological control of the blood FFA levels. We hypothesized that glyceroneogenesis could be a ROR␣ target in adipose tissue for the following reasons. 1) This pathway shares with liver gluconeogenesis the same key gene, PCK1; 2) previously, we revealed that GOT1, encoding aspartate aminotransferase, another enzyme involved in glyceroneogenesis, was an in vitro ROR␣ target gene (45); and 3) RORsg/sg mouse, a natural mutant ROR␣-deficient strain (13), exhibits a lean phenotype with decreased fat mass and small adipocytes for an unsolved reason (26). Indeed, adipocyte precursors from these mice are fully competent to differentiate in vitro (12, 35). This raised the possibility that RORsg/sg mice could harbor specific adipose defects that limit their capacity for energy storage. We reasoned that these defects could result from altered glyceroneogenesis because their lean phenotype is reminiscent of that seen in adipose-specific PCK1-invalidated mice. Thus, our aim was to investigate whether glyceroneogenesis was a target pathway of ROR␣ in adipose tissue, as proposed for gluconeogenesis in the liver. We investigated both pathways in RORsg/sg mice by comparison with their wild-type littermates. We also performed ex vivo studies by using ROR␣ agonists and antagonists on liver and adipose tissue explants. Our data clearly showed that these two important metabolic pathways were impaired in RORsg/sg mice concomitantly with an altered expression and activity of PCK1. The present study demonstrated the physiological function of ROR␣ in regulating both glucose and lipid homeostasis. MATERIALS AND METHODS

Animals and experimental design. Animal studies were conducted according to the French guidelines of the Charles Darwin Ethics Committee (Ce5/2010/034) and were approved by this committee. All mice were maintained in a specific pathogen-free environment. The staggerer mutation was maintained in a C57BL6J genetic background in our breeding colony. It was developed from mice kindly provided by Prof. Jean Mariani (UMR Neurobiologie des Processus Adaptatifs, UPMC, Paris, France). Staggerer (RORsg/sg) and wild-type (ROR⫹/⫹) mice were obtained by crossing fertile heterozygous (RORsg/⫹) mice and identifying homozygous offspring by PCR genotyping. RORsg/sg mice have decreased ROR␣ expression in all tissues (20, 26) and a phenotype similar to the engineered ROR␣-KO mice (42). The truncated ROR␣ protein was unstable and not detectable (5, 13). Mice were housed at 24°C in a temperature-controlled room with a 12-h light-dark cycle. Water and food (A03; UAR, Epinay-sur-Orge, France) were provided ad libitum. A03 food was provided mashed and moistened for RORsg/sg mice because their ataxia prevents them from feeding normally. RORsg/sg mice were housed in separate cages with their mother. Only 4- to 6-mo-old male mice were used in this study. The pyruvate tolerance test was performed by intraperitoneally injecting 16-h-fasted mice with a 2 g/kg buffered pyruvate solution.

Blood glucose levels were then checked every 15 min for 2 h. For the analysis of the rhythmic pattern of protein expression as well as of FFA release into the blood, mice were euthanized by cervical disruption every 3 h after the beginning of the fasting period [with food being removed at Zeitgeber time 0 (ZT0)]. The food was reintroduced into the cage after 10 h (ZT10). However, we noticed that RORsg/sg but not WT mice were still asleep at ZT10, which have delayed their refeeding by about 2 h. Functional analysis of glyceroneogenesis in adipose tissue explants. Inguinal adipose tissue from 4-h-fasted mice was minced into very small pieces (5–10 mg) and incubated in DMEM without glucose containing 3% BSA for 4 h and then in Krebs-Ringer bicarbonate buffer containing 3% fatty acid-free BSA, [14C1]pyruvate (2 ␮Ci/ml), 0.5 mM pyruvate, and 1 ␮M isoproterenol. After 1 h, the incubation medium was sampled for the estimation of lipolytic FFA and glycerol. The corresponding tissue explants were frozen in liquid nitrogen before lipid extraction by the simplified method of Bligh and Dyer (4). The subsequent incorporation of [14C1]pyruvate into the lipid moiety was estimated by counting the radioactivity associated with this fraction and was used to appreciate the level of FFA that had been reesterified during the 1-h lipolytic process. For the study of glycerol incorporation into triglycerides, reflecting glycerol/kinase activity, [14C]glycerol (2 ␮Ci/ml) and 0.5 mM glycerol were added to the Krebs-Ringer buffer, and the experiment was performed without isoproterenol. Glucose output from liver slices. Precision-cut liver slices from 4-h-fasted mice were prepared with a tissue chopper (10), preincubated in DMEM containing 0.5% BSA (without glucose) for 2 h, and then incubated in Krebs-Henseleit bicarbonate buffer containing 0.5% BSA, 1 mM lactate, and 0.1 mM pyruvate for 1 h at 37°C with constant shaking. The incubation medium was then assayed for glucose concentration (GO Assay Kit; Sigma-Aldrich, L’isle d’Abeau-Chesnes, France). To examine the pyruvate-dependent glucose production that comes from gluconeogenesis, the liver slices were preincubated 6 h before performing the glucose production test in the absence or presence of 10 ␮M ROR ligands or excipient. Western blotting. Frozen adipose tissue was homogenized on ice in Tissue Protein Lysis Buffer (Euromedex, Souffelweyersheim, France) and then processed as described in Leroyer et al. (31). Because of liver zonation, the entire liver was first pulverized under liquid nitrogen with a pestle and mortar and mixed. Equal amounts of protein were pooled from two mice at the indicated time points. A total of 20 ␮g of protein extract was used for immunoblot analysis. For immunoblot detection, anti-PCK1 (My BioSource; CellGenetech, Paris, France) and anti-G6PC (SC-25840; Santa Cruz Biotechnology, Heidelberg, Germany) antibodies were used. RNA extraction and quantitative RT-PCR. Total RNA was extracted using the RNeasy mini kit (Quiagen, Courtaboeuf, France) from snap-frozen tissue in liquid nitrogen after euthanasia and then homogenized on ice. cDNA synthesis from 1 ␮g of RNA was performed using High Capacity cDNA Reverse Transcription kit from Applied Biosystems (Carlsbad, CA) and quantitative reverse transcription-polymerase chain reaction with 10 ng of cDNA and 500 nM primers, using SYBR Green and a Roche thermocycler (Roche Diagnostics, Meylan, France). Gene expression was normalized with two housekeeping genes (GAPDH and 36B4). Data analysis was based on the ⌬⌬CT method. Primer sequences were as follows: for mPCK1, CAACTTCGGCAAATACCTG (forward) and CTGTCTTCCCCTTCAATCC (reverse); for mG6PC, ATGAACATTCTCCATACTTTGGG (forward) and GACAGGGAACTGCTTTATTATAGG (reverse). Blood and tissue biochemical analysis. Blood glucose was measured using a glucometer (Roche Diagnostics, Meylan, France) on a drop from the tail. Plasma insulin and glucagon were determined using the Milliplex kit (Millipore, St. Quentin en Yvelines, France) on a BioPlex 200 system (Bio-Rad, Marnes-la-Coquette, France). Glycerol and FFA from culture media were quantified using kits according

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

E107

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

Metabolic phenotype of RORsg/sg and WT mice. The average body weight of 6-mo-old male RORsg/sg mice was 25% lower than that of their wild-type (WT) littermates (Fig. 1A). This lean phenotype was characterized by a 50% decreased fat mass, affecting both epidydimal (EAT) and inguinal subcutaneous adipose tissue (SAT), whereas the weight of the liver relative to that of the whole body was unchanged (Fig. 1A). After a 6-h fast, RORsg/sg mice were hypoglycemic without WT

30

*

20 10 0

WT

B

6

Blood glucose (g/l)

sg/sg

(relative tissue weight)

Body weight (g)

A

5 4 3 2

*

*

1 0 EAT

SAT

liver

sg/sg

2.5 2

*

1.5 1 0.5 0

Plasma glucagon (ng/l)

RESULTS

variation of insulin and glucagon levels compared with their WT littermates (Fig. 1B). To explore in vivo a potential defect of glyceroneogenesis in RORsg/sg mice that would impact on metabolic parameters, we measured the release of FFA from adipose tissue into the blood during their fasting period (with mice being housed in cages without food when the light turns from ZT0 to ZT10; Fig. 1C). During this fasting period the plasmatic FFA level of WT mice rose to a peak occurring between 3 and 6 h, with a progressive decrease thereafter, as shown by others (22, 41). This plasmatic FFA increase is the consequence of the lipolysis of TG stored in AT. The decrease is due to both FFA utilization as energy source and reesterification in AT as a way to adjust their level according to body needs. In RORsg/sg mice, blood FFA levels exhibited a delayed peak at 9 h postfasting, with a higher FFA level than that seen in the WT mice before decrease (Fig. 1C). This delayed peak could reflect impaired FFA reesterification via glyceroneogenesis, which usually takes place after several hours of fasting (40), whereas the important decrease at ZT12 could be related to the enhanced FFA oxidation described in these mice (20, 28). To further investigate in vivo a potential defect of gluconeogenesis in RORsg/sg mice, we performed a PTT, which consisted of analyzing the rise of glycemia in response to a pyruvate injection in a fasted subject. A clear increase in

Plasma insulin (ng/l)

to the manufacturer’s instructions (Free-glycerol reagent from SigmaAldrich and free fatty acids, half-microtest from Roche Diagnostics). Specific activity of the cystosolic form of PEPCK (PEPCKc) was spectrophotometrically measured on postmitochondrial supernatant isolated from adipose tissue, as mentioned previously (44). ROR␣ and Rev-erb␣ chemical ligands. 7␤-Hydroxycholesterol and GSK412 were purchased from Sigma Aldrich, SR1078 was purchased from Calbiochem (Merck Chemicals, Nottingham, UK), and SR1001 was purchased from Cayman (Interchim, Montigny le Bretonneux, France). Statistical analysis. Values are means ⫾ SD of the indicated number of measurements. Student’s t-test was used when only two groups were compared; one-way ANOVA was used when more than two groups were compared and when only one parameter was investigated, followed by post hoc analysis. P ⬍ 0.05 was considered the limit for statistical significance.

1

0

60 40 20 0

D

C

Blood glucose (g/l)

1000

FFA μmol/l

800 600

* 400 200

1.5 3.5

1.0

3

0.5

2.5

0.0

*

WT sg 2

1

*

*

1.5

*

0.5

Fasting time (h)

0 0

3

0

6

hours

9

12

T0

T30

T60

T90

T120

minutes post-pyruvate

Fig. 1. Metabolic phenotype of retinoid-related orphan receptors (ROR)sg/sg mice. Male 4- to 6-mo-old mice were used. RORsg/sg (sg/sg) were compared with their ROR⫹/⫹ littermates [wild type (WT)]. Values are expressed as means ⫾ SD and compared with Student’s t-test. *P ⬍ 0.05 vs. WT genotype. Open bars and 䊐, WT mice; black bars or , sg/sg mice. A: body and tissue weights. EAT, epididymal adipose tissue; SAT, inguinal subcutaneous adipose tissue (n ⫽ 18/group). B: glycemia, insulinemia, and glucagonemia of 6-h-fasted mice (n ⫽ 18/group). C: rise of plasmatic free fatty acids (FFA) of mice measured during the fasting period. Mice were transferred to a food-free cage at ZT0 (at the beginning of the inactive period), and blood samples were collected every 3 h (n ⫽ 6 –9/group). Food was proposed at ZT10. D: pyruvate tolerance test performed in overnight fasted mice (n ⫽ 4/group). Hepatic glucose production (HGP) was evaluated in vivo by assaying the glucose production capacity after a pyruvate bolus injection (2 g/kg). Inset: values of glycemia area under the curve (AUC) expressed relative to WT mice. AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

E108

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

Our data show that the incorporation of pyruvate into TG was decreased by one-half in the adipose tissue explants from RORsg/sg mice compared with WT mice (Fig. 2A). In parallel, the FFA/glycerol ratio measured in the corresponding culture medium was increased from 2 to about 3 (Fig. 2B). This was related to the increase in FFA because the glycerol release was not different between the two genotypes (1.66 ⫾ 0.38 and 1.95 ⫾ 0.41 nM·h⫺1·g⫺1 for WT and RORsg/sg mice, respectively). We also investigated a potential involvement of glycerolkinase (GyK) in FFA reesterification by providing [14C]glycerol as the precursor of G3P in SAT explants from mice fasted for 6 h. Our preliminary data did not show a significant difference between the two genotypes (3.70 ⫾ 0.74 and 4.62 ⫾ 0.55 nM·h⫺1·g⫺1 of WT and RORsg/sg mice, respectively, n ⫽ 4) but would require additional experiments. Thus, our data illustrated a functional defect of FFA reesterification via glyceroneogenesis in SAT in the absence of ROR␣ that could at least partly explain the decreased size of the adipocytes from adult RORsg/sg compared with WT mice (Fig. 2F). This was further assessed by the observation that the PCK1 mRNA amount, the PEPCKc specific activity, and PEPCKc protein were decreased in SAT of RORsg/sg mice compared with those of WT mice fasted for 6 to 9 h (Fig. 2, C–E). Moreover, Fig. 2E showed that, at the protein level, PEPCKc expression was not only diminished but also delayed; this could be related to the delayed food intake of RORsg/sg mice (observed around ZT12; indeed they were still sleeping at the time the food was reintroduced). Finaly, glyceroneogenesis was also found to be altered in the other adipose depots of RORsg/sg mice fasted for 6 h, with PCK1 mRNA amount being decreased by 50% in epidydimal AT, by 65% in perirenal AT,

glucose level was observed in WT animals, with a maximal increase of 80% observed after 60 –90 min. By contrast, after an overnight fast, the level of blood glucose in RORsg/sg mice was already reduced by 33% compared with WT mice and then stayed significantly lower after the pyruvate bolus (Fig. 1D). The corresponding glycemia area under the curve (AUC) value in RORsg/sg mice was only the third of that observed in WT mice (shown in Fig. 1D, inset). These results clearly indicated a defect in glucose production from pyruvate in RORsg/sg compared with WT mice. Glyceroneogenesis is altered in adipose tissue of RORsg/sg mice and modified by ROR␣ ligands. To investigate whether the higher FFA plasmatic level observed in 9-h-fasted RORsg/sg mice compared with WT littermate could be the result of a defect of their reesterification via glyceroneogenesis, we analyzed this pathway in adipose tissue explants issued from mice of the two genotypes. Explants were isolated from 4-h-fasted mice and incubated in glucose, pyruvate, and insulin-free medium for 4 h. Then, the functional appreciation of glyceroneogenesis was performed for 1 h under an acute lipolytic situation in the presence of [14C1]pyruvate. [14C1]pyruvate was used because, in contrast to C2- or C3-labeled molecules, only the C1 carbon of pyruvate is conserved in the G3P moiety of newly synthesized TG and is thus a marker for glyceroneogenesis. Under these conditions, the proportion of the FFA produced by lipolysis and then reesterified back into the adipocytes can be measured because they are incorporated as 14 C1-labeled G3P in TG. In the culture medium, the FFA/ glycerol ratio, when lower than 3, is used as a glyceroneogenic index (when equal to 3, this means that no reesterification has occurred; indeed, there are 3 FFA linked to 1 glycerol in a TG molecule).

A

B

Pyruvate incorporation into TG

PEPCK-c activity

4

6

*

10

2 1

5 0

0

1

nM/mn/mg prot

*

3

15

relative mRNA level

20

sg/sg

(nM/g)

D

PCK1 mRNA

1.5

25 WT

C

FFA/Glycerol ratio

* 0.5

0

E

4

**

2

0

F

PEPCKc

actin ZT

3

6

9

WT

12

15

18

3

6

9

12

15

18

sg/sg

wt sg/sg

sg/sg

Fig. 2. FFA reesterification via glyceroneogenesis is altered in adipose tissue from ROR mice compared with WT littermates. Inguinal fat pads were isolated from 4-h-fasted mice and then tested for glyceroneogenesis, as described in MATERIALS AND METHODS, or used for RNA and protein extraction (n ⫽ 10/group). Values are expressed as means ⫾ SD and compared with Student’s t-test. *P ⬍ 0.05 and **P ⬍ 0.02 vs. WT genotype. A: radiolabeled pyruvate incorporation evaluates the amount of FFA reesterified back into TG during a 1-h lipolytic period in 4-h-fasted mice. B: FFA release in the culture media expressed vs. glycerol release. FFA/glycerol values correspond to the means of the ratios determined in each animal. Glycerol release was 1.66 ⫾ 0.38 and 1.95 ⫾ 0.41 nM·h⫺1·g⫺1 AT from 4-h-fasted WT and sg/sg mice, respectively. C: mRNA expression level of PCK1 in 6h-fasted mice D: phosphoenolpyruvate carboxykinase c (PEPCKc)-specific activity in 6-h-fasted mice. E: Western blot analysis of the rhythmic profile of PEPCKc protein according to the feeding status. Mice were transferred to a food-free cage at Zeitgeber time 0 (ZT0; as indicated by the dashed gray arrow) and then refed from ZT10 to ZT18 (solid black arrow). However, RORsg/sg mice slept longer than WT mice, thus delaying their food intake after ZT10. Equal amounts of protein were pooled from 2 WT and sg/sg mice at the indicated time points. Actin is used as a control of depot. F: photomicrographs of WT and RORsg/sg mice adipose tissue (similar scale). AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

and by 81% in brown AT compared with their similar depots in WT littermates (data not shown). We then tested the effects of ROR␣ antagonist and agonist on glyceroneogenesis in WT and RORsg/sg mice (Fig. 3). SAT explants were preincubated overnight with 10 ␮M ROR␣ ligands (to fit the acrophase of circadian expression) and then tested for glyceroneogenesis, as presented previously. In WT mice, 7␤-hydroxycholesterol, a ROR␣ antagonist (47), decreased the incorporation of pyruvate into TG and simultaneously increased the FFA/glycerol release in the medium. On the contrary, SR1078, a ROR␣ agonist (46), did the opposite. It had no effect on the adipose tissue of RORsg/sg mice, thus confirming the specific involvement of ROR␣ in the modulation of FFA homeostasis via glyceroneogenesis in adipose tissue. Because Rev-erb␣ is a physiological inhibitor of the binding of ROR␣ to RORE sequences, we also tested a synthetic Rev-erb␣ agonist (GSK412) that was shown to repress PCK1 expression in HepG2 cells (16). To that end, adipose tissue explants were treated during the day period (to fit the acrophase of Rev-erb␣ circadian expression). GSK412 exerted an inhibitory effect on glyceroneogenesis and increased FFA release, as observed with the ROR␣ antagonist, suggesting that Rev-

fold vs DMSO

A

Pyruvate incorporation into TG 2

*

1.5

1

*

*

*

*

0.5

0 DMSO 7βOH SR1078 GSK

B

FFA/Glycerol 2.5

sg/sg

wt

*

*

2

fold vs DMSO

DMSO SR1078

* *

1.5 1

*

0.5 0 DMSO 7βOH SR1078 GSK

wt

DMSO SR1078

sg/sg

Fig. 3. Glyceroneogenesis is modulated by ROR␣ ligands in adipose tissue. Inguinal fat pads were isolated from 4-h-fasted mice of each genotype, preincubated overnight with 10 ␮M ROR␣ ligands (or daily with 10 ␮M Rev-erb␣ ligand), and then tested for glyceroneogenesis, as described in MATERIALS AND METHODS (n ⫽ 6 – 8/group). A: measurement of pyruvate incorporation into adipose tissue TG, which reflects the glyceroneogenicdependent FFA reesterification. B: FFA release in the corresponding culture media expressed vs. glycerol release. Values are expressed as means ⫾ SD and compared with 1-way ANOVA. *P ⬍ 0.05 vs. WT genotype.

E109

erb␣ would act as an inhibitor of FFA reesterification via glyceroneogenesis, whereas ROR␣ would act as an activator. Hepatic gluconeogenesis is altered in RORsg/sg mice and modified by ROR ligands. To confirm our in vivo observations showing a blunted response of glycemia upon pyruvate injection in RORsg/sg mice, we compared the ex vivo glucose production in liver slices from RORsg/sg and WT mice. The first set of experiments, which were performed in freshly isolated liver slices from 4-h-fasted mice, showed a 56% decrease in glucose released by the RORsg/sg liver slices compared with their WT littermates (Fig. 4A). This decreased production was found to be independent from the presence of lactate-pyruvate (LP) in the medium. Therefore, we postulated that it likely reflected a defect of glycogenolysis in RORsg/sg mice. Indeed, during glycogenolysis, which is the first pathway activated in the fasting condition to provide hepatic glucose production (HGP), there is no requirement for an activation of PCK1, which is limited to the process of gluconeogenesis (7). However, glycogenolysis needs the presence of glucose-6phosphatase, which controls the last step of glucose production and whose encoding gene G6PC is a target of ROR␣ (7). Accordingly, we confirmed a major decrease in G6PC mRNA amount in the liver of RORsg/sg mice compared with WT mice fasted for 6 h (Fig. 4C). Moreover, the study of the protein expression according to the fasting/fed status confirmed a lower amount of glucose-6-phosphatase in fasted RORsg/sg than in WT mice (Fig. 4D). In a second set of experiments, we measured glucose production from liver slices that had been incubated for 6 h in the absence of insulin to allow the transition to gluconeogenesis. Under these conditions, we were able to observe a LP-dependent glucose production in liver slices of WT mice, reflecting gluconeogenic activity, that did not occur in liver slices of RORsg/sg mice (Fig. 4B). Accordingly, the amount of PCK1 mRNA was decreased by half in liver slices from fasted RORsg/sg compared with those from WT mice (Fig. 4C). As expected, in the liver of WT mice, the amount of the PEPCKc protein increased throughout the fasting period and decreased upon refeeding (2 and 5 h after feeding) (Fig. 4D). Interestingly, however, the increase in the PEPCKc protein was compromised in the RORsg/sg mice, especially between 6 and 9 h of the fasting period, before increasing at ZT12 and then decreasing at ZT15. The increase observed at ZT12 is probably related to a longer fasting period in these mice and confirmed a delayed response to fasting of PCK1 in the absence of ROR␣ but a conserved inhibition by refeeding. We completed the investigation by measuring the enzyme-specific activity. Data confirmed that hepatic PEPCKc activity was altered in the absence of ROR␣ in both 9-h-fasted mice (at ZT9), and 6-h-refed mice (at ZT15). In a third set of experiments, we tested the effects of a ROR␣ antagonist and agonist on glucose production in liver slices issued from WT and RORsg/sg mice (Fig. 5). Liver slices were preincubated for 6 h with 10 ␮M ROR␣ ligands and then tested in the presence of LP for 2 h. Data obtained with slices from WT mice showed a mild but significant decrease in glucose production when liver slices were treated with the antagonist (7␤-hydroxycholesterol), whereas an increased glucose production occurred when treated with the agonist (SR1078). These effects were not observed in liver slices from RORsg/sg mice (Fig. 5A). We further verified the effect on

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

E110

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

6

*

4

*

2 0 LP

-

+

-

wt

+

sg/sg

relative mRNA level

Glucose production mg/g

Glucose production mg/g

8

C #

0.6

0.4

0.2

0.0

-

LP

+

-

1.2 1 0.8 0.6

D

***

0.4 0.2

1.5

*

1

0.5

0

0 wt sg/sg

+

sg/sg

wt

PCK1

G6PC

1.4

relative mNA level

B

A

E

wt sg/sg

PEPCKc activity

PEPCKc G6Pase actin ZT

0

3

6

9

12

15

0

3

6

9

sg/sg

WT

12

15

nM/mn/mg protein

14 12 10 8 6

*

wt sg/sg

4

*

2 0

Fasted

Refed

sg/sg

Fig. 4. A and B: glucose production is altered in liver slices from ROR compared with WT mice. Glucose production by liver slices from RORsg/sg and WT mice incubated with (⫹) or without (⫺) lactate-pyruvate (LP) for 1 h in freshly prepared liver slices from 4-h-fasted mice (A) and in liver slices that have been preincubated for 6 h in the absence of insulin to deplete glycogen stores (B) (n ⫽ 6/group). C: mRNA expression levels of G6PC and PCK1 in liver slices obtained from 6-h-fasted mice (n ⫽ 10/group). D: Western blot analysis of the rhythmic profile of hepatic PEPCKc and glucose-6-phosphatase (G6pase) protein expression according to the feeding status. Mice were fasted at the beginning of the light period (ZT0; gray arrow) and refed from ZT10 (black arrow). Sg/sg mice slept longer than WT mice and were refed later than ZT10. Equal amounts of protein were pooled from 2 WT and sg/sg mice at the indicated time points. Actin was used as a control of depot. E: PEPCKc-specific activity in the liver of 9-h-fasted mice (at ZT9), which were then refed (at ZT15) (n ⫽ 4). Values are expressed as means ⫾ SD and compared with Student’s t-test. *P ⬍ 0.05 and ***P ⬍ 0.01 vs. WT genotype; #P ⬍ 0.05 vs. without LP.

PCK1 gene expression of the treatment of liver slices from WT mice with two different ROR␣ antagonists. Figure 5B illustrates a decrease in hepatic PCK1 mRNA amount induced by both 7␤-hydroxycholesterol and SR1001, thus confirming their inhibitory effects. DISCUSSION

In this study, we demonstrated the involvement of ROR␣ in the positive control of two close metabolic pathways, gluconeogenesis in the liver and glyceroneogenesis in adipose tissue, both of which are dependent on a common key gene, PCK1. Both are required, under fasting situations, to produce glucose and G3P from lactate and pyruvate, respectively. The first one is involved in glucose homeostasis and the second one in lipid homeostasis. We showed that, in the absence of ROR␣, mice presented both impaired HGP and defective FFA storage in their fat depots. These in vivo physiological observations were further assessed by ex vivo functional analysis of the two pathways in tissue explants. They confirm the role of ROR␣ in regulating liver gluconeogenesis and demonstrate for the first time its role in controlling FFA homeostasis in white adipose tissue via glyceroneogenesis and thus fat mass. Previous observations performed by using the RORsg/sg mouse provided insight into the role of ROR␣ in metabolic homeostasis. These ROR␣-deficient mice exhibited a fat mass reduced by half and fasting hypoglycemia for unresolved reasons. The fact that Rev-erb␣-KO mice display the opposite phenotype (increased adiposity and fasting hyperglycemia) suggested that ROR␣ and Rev-erb␣ could cross-talk as transactivator and repressor, respectively, on common pathways in both glucose and lipid metabolism.

We first attempted to confirm that ROR␣ acts (as a physiological actor) on gluconeogenesis in vivo, as suggested by Kumar et al. (24). The induction of hepatic gluconeogenesis during fasting periods is an essential adaptive response that requires the enhanced transcription of genes encoding the rate-determining gluconeogenic enzymes PCK1 and G6PC that we confirmed at the protein level in WT mice. It has been well established that the transcription of the PCK1 gene was tightly regulated by diet-dependent hormones (insulin and glucagon) (see Ref. 48 for review). However, the decreased PCK1 mRNA amount we observed in the liver of RORsg/sg mice fasted for 6 h by comparison with WT mice was likely due not to abnormal levels of these hormones (indeed, their plasmatic level did not differ between the two genotypes) but rather to the absence of ROR␣. This result is different from that reported in a previous study showing that liver PCK1 mRNA amount did not change between RORsg/sg and WT mice (19). This discrepancy can be explained by the fact that they obtained these data on mice in the feeding period (ZT19), but with no control of the exact time of food intake. Indeed, PCK1 transcription is acutely regulated by diet (1), and the measurement of its expression requires strict feeding control. Furthermore, we confirmed an altered response of the PEPCKc to fasting at the protein level in the liver of RORsg/sg mice. Thus, our data confirm the positive role of ROR␣ in regulating hepatic gluconeogenesis (24), possibly by controlling PCK1 expression in addition to G6PC (7), and suggest that it could participate with Rev-erb␣ to a circadian control of hepatic glucose production. The expression of ROR␣ and Rev-erb␣ oscillates in a circadian manner (12 h out of phase with one another), with

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

A Glucose production (mg/g)

0.7

##

0.6 0.5 0.4 0.3

#

0.2 0.1 0

DMSO 7βOH SR1078

DMSO 7βOH SR1078

wt

sg/sg

B 1.4

PCK1

relative mRNA level

1.2 1.0 0.8 0.6 0.4

### ###

0.2 0.0

DMSO 7βOH SR1001

Fig. 5. Hepatic gluconeogenesis is modified by ROR␣ ligands. A: glucose production by liver slices from RORsg/sg and WT mice preincubated overnight with ROR␣ ligands (10 ␮M) or excipient (DMSO) and then incubated with LP for 1 h. B: effect of 2 different ROR␣ antagonists on PCK1 mRNA expression level in overnight-treated liver slices. Values are expressed as means ⫾ SD and compared with 1-way ANOVA. #P ⬍ 0.05, ##P ⬍ 0.05, and ###P ⬍ 0.01 vs. DMSO.

Rev-erb␣ gene expression peaking during the sleeping period and ROR␣ during the awake period (see Ref. 2 for review). However, the rhythmicity of ROR␣ in the liver is rather weak compared with that of Rev-erb␣ (22). Thus, in such a scenario, one can speculate that ROR␣ would be more implicated in the basal transcription of PCK1 that would be rhythmically interrupted by the brief peak of Rev-erb␣ protein (that could compete for RORE binding at ZT10, when its expression is maximal). Indeed, it has been shown that the circardian expression of Rev-erb␣ drives its rhythmic binding genome wide (15). In the RORsg/sg mice, the rhythmic expression pattern of Rev-erb␣ was shown not to be altered in the liver (43). Thus, hypoglycemia observed in these mice could be due to the absence of a transactivator (ROR␣) rather than to the overexpression of an inhibitor (Rev-erb␣). On the contrary, the absence of an inhibitor of gluconeogenesis (Rev-erb␣) should lead to glucose overproduction. However, the hyperglycemia observed in the Rev-erb␣ knockout mice was not attributed to an increased gluconeogenesis (9). We suspect that this discrepancy with the data of Yin et al. (50) could come from the fact that the PTT was performed at ZT2 [at the very beginning of the inactive period, i.e., at a period when Rev-erb␣ protein is nearly absent in WT mice (38)]. We

E111

think that to be meaningful this analysis should have been performed around ZT12 (at the end of the inactive period, after the physiological peak of Rev-erb␣ protein) to be able to compare a mouse expressing Rev-erb␣ with a Rev-erb␣⫺/⫺ mouse. We then showed that ROR␣ controlled glyceroneogenesis in adipose tissue and thus could be a regulator of lipid storage in fat. We reasoned that the reduced adiposity of RORsg/sg mice, being associated with a smaller adipocyte size (Fig. 2F and Ref. 26), could be due to a diminished deposition of TG in adipose tissue under fasting conditions. Indeed, reduced adiposity was not described to be linked to a defect of adipogenesis, which, on the contrary, was found to be improved (12, 35), nor to an altered lipolysis [observation based on the gene expression of two major lipolytic enzymes, hormone-sensitive lipase (HSL) and adipose triglyceride lipase (ATGL) (26)]. We confirmed that RORsg/sg mice exhibited no clear defect in lipogenesis under feeding conditions on the basis of similar sterol regulatory element-binding protein-1c (SREBP)-1c, PPAR␥, or fatty acid-binding protein 4 mRNA amount in inguinal fat (not shown). Moreover, preliminary data did not suggest any significant variation in perilipin protein expression between the two genotypes. The lean phenotype of RORsg/sg mice was reminiscent of the adipose tissue-specific PCK1 invalidated mice model, which showed that the ablation of glyceroneogenesis in adipose tissue, and therefore, the inhibition of TG synthesis in fasting periods, was responsible for the reduced fat mass in these lipodystrophic mice (36). These results have clearly established the physiological role of glyceroneogenesis in maintaining fat homeostasis in adipose tissue. Accordingly, our study also described a defect of glyceroneogenesis in each fat depot that was associated with a reduced fat mass in the RORsg/sg mice and suggested a new possible explanation for the lean phenotype of these mice. Another cause was also recently attributed to an induced expression of the uncoupling protein 1 (UCP1) in their fat (28); accordingly, we also observed that UCP1 expression was increased in SAT from RORsg/sg mice compared with WT mice (not shown). Indeed, their lean phenotype was not due to a decreased energy intake because there was no lower food intake nor impaired lipid intestinal absorption in these mice (Ref. 20 and Kadiri S and Antoine B, unpublished observation). We also analyzed the involvement of GyK in FFA reesterification in inguinal fat; it was found to be weak and not significantly different between the two genotypes. However, given the small number of tested animals, these data will require additional experiments. Furthermore, we cannot exclude a potential increase in pyruvate dehydrogenase (PDH) activity in RORsg/sg mice that could increase the decarboxylation of pyruvate to acetyl-CoA and thus decrease the C1pyruvate availability for glyceroneogenesis. Nevertheless, high levels of acetyl-CoA resulted in the inhibition of pyruvate decarboxylation by suppressing PDH, making more pyruvate available for glyceroneogenesis (21). Furthermore, our demonstration of a common decrease in PEPCKc at the level of mRNA, protein, and specific activity in RORsg/sg mice consolidates our functional data based on pyruvate incorporation into TG. Thus, our data identified a new couple of positive (ROR␣) and negative (Rev-erb␣) regulators of glyceroneogenesis in

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

E112

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

adipose tissue and potentially of fat homeostasis. To our knowledge, we describe here for the first time both an induction of glyceroneogenesis by a ROR␣ agonist and its inhibition by a ROR␣ antagonist or by a Rev-erb␣ agonist. Such a cross-talk between ROR␣ and Rev-erb␣ could allow for some circadian control of fat mass via the control of glyceroneogenesis. Accordingly, it was shown recently that the in vivo modulation of Rev-erb␣ activity by a synthetic agonist was able to induce a loss of fat mass in obese mice. Finally, the rhythmicity of ROR␣ expression is higher in adipose tissue than in the liver (49), which could explain the more pronounced phenotype seen in adipose tissue (fat pad weight) compared with the liver (glycemia) when ROR␣ is missing. We further suggested that ROR␣ could act as a regulator of FFA homeostasis via glyceroneogenesis. An important function of adipose tissue is to store excess energy substrate in the form of lipids during the fed state and to release those stored lipids to be used under fasting conditions. The timing of FFA release from adipose stores has to be tightly controlled, as excess of circulating FFA may lead to lipotoxicity and promote diabetic disorders. Some circadian regulation of lipid homeostasis was demonstrated recently in white adipose tissue. For example, clock genes (Clock and Bmal1) regulate rhythmic FFA release by driving rhythmic transcription of two genes encoding the lipolysis pacemaker enzymes, ATGL and HSL (41). In this study, we show that the ROR␣-deficient mice exhibited both an increased FFA release into the blood in the late fasting period, at a time where glyceroneogenesis should have taken place to limit their release, and a decreased fat mass. On the contrary, Rev-erb␣-KO mice present the inverse phenotype (a reduced FFA availability when fasted accompanied with increased body fat) (9). The comparison of these two genetic models suggests that the regulation of glyceroneogenesis by ROR␣ and Rev-erb␣ could also impact FFA homeostasis. It will be interesting to check whether the phenotype of Rev-erb␣ knockout mice could originate at least in part from increased glyceroneogenesis, that is to say increased FFA reesterification in their adipose tissue, which could limit FFA release into the circulation. Finally, concerning the role of liver in FFA homeostasis, we observed a decreased mRNA amount of SREBP-1c and fatty acid synthase in the liver of 6-h-fasted RORsg/sg mice compared with WT mice (data not shown), as found previously (26). Reduced hepatic de novo lipogenesis, as a result of reduced SREBP-1c and FAS, could indeed participate in reduced VLDL production and storage of TG inside adipocytes, thereby reducing adiposity. Regarding liver glyceroneogenesis, it was estimated by Martins-Santos et al. (33) that it would contribute to only 20% of the G3P generated for TG synthesis in the liver of fasted rats. To evaluate that point in our mice, we compared pyruvate incorporation into the liver of fasted WT and RORsg/sg mice (n ⫽ 4). We found that its level, although low, was significantly decreased (⫻0.67) in the liver of RORsg/sg mice (4.22 ⫾ 0.20 nM·h⫺1·g⫺1) compared with WT mice (6.15 ⫾ 0.59 nM·h⫺1·g⫺1). Thus, our data suggested that glyceroneogenesis could also play a role in the lower hepatic TG content of RORsg/sg mice on a chow diet. However, provided the small number of tested animals, confirmation will require more experiments. Taken together, our data confirm the control of hepatic gluconeogenesis by ROR␣ and identify ROR␣ and Rev-erb␣

as a new loop of circadian control of glyceroneogenesis in adipose tissue. We suggest that these two pathways could be activated by ROR␣ during the active period, thus possibly allowing the body to anticipate for glucose (or G3P) production during the inactive (and fasting) period, then repressed by Rev-erb␣ at the end of the inactive period. Such a regulation could be rather independent of diet-dependent hormone levels, as observed here. Indeed, it is well known that there are interactions between circadian clock and metabolism, and there is a need to further examine the function of ROR␣ (and Rev-erb␣) in specific tissues. Further molecular studies are also required to find the involved RORE sequence on the PCK1 promoter to be certain of a direct control. Indeed, one cannot exclude that the observed effects of ROR␣ and Rev-erb␣ on both glucose and FFA metabolism via PCK1 could have been mediated by some of their other target genes, such as the clock-genes Bmal1 or Clock. In the liver, PCK1 was shown to be a cis-target of both Rev-erb␣ and Bmal1, whose binding sites overlap in the same “cistrome” (6). RORs are constitutively active but also function as liganddependent transcription factors, thus being considered as potential therapeutic targets for human diseases. This prompted the development of synthetic ligands (see Ref. 23 for review). ROR␣-deficient mice display ataxia and cerebellar atrophy due to an altered maturation of Purkinje cells at the fetal and postbirth periods (18). However, the simultaneous absence of ROR␣ in several metabolic tissues such as the liver, adipose tissue, and muscle resulted in a globally metabolically favorable phenotype in the adult mice; indeed, the RORsg/sg mice exhibited decreased cholesterol synthesis (32, 39), reduced lipogenesis (20, 26), and gluconeogenesis (this study) in the liver and decreased glyceroneogenesis in adipose tissue (this study) associated with increased energy expenditure in the muscle (20, 22), SAT, and brown adipose tissue (28). This highlights the importance of this clock gene in regulating glucose and lipid metabolism and could help designing new ways of fighting against metabolic disorders commonly seen in type 2 diabetes. ACKNOWLEDGMENTS We thank Laetitia Dinard for kind and trustable help in mice breeding and Marie Garcia for mice genotyping. GRANTS S. Kadiri was the recipient of a Ph.D grant from the French Ministère de l’ Education Nationale et de la Recherche. B. Antoine is a CNRS researcher. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the authors. AUTHOR CONTRIBUTIONS S.K., C.M., M.G., T.L., and B.A. performed experiments; S.K. and B.A. analyzed data; S.K. and B.A. prepared figures; S.K., C.M., M.G., T.L., J.C., and B.A. approved final version of manuscript; J.C. edited and revised manuscript; B.A. conception and design of research; B.A. interpreted results of experiments; B.A. drafted manuscript. REFERENCES 1. Antras-Ferry J, Robin P, Robin D, Forest C. Fatty acids and fibrates are potent inducers of transcription of the phosphoenolpyruvate carboxykinase gene in adipocytes. Eur J Biochem 234: 390 –396, 1995.

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS 2. Bass J, Takahashi JS. Circadian integration of metabolism and energetics. Science 330: 1349 –1354, 2010. 3. Beale EG, Hammer RE, Antoine B, Forest C. Glyceroneogenesis comes of age. FASEB J 16: 1695–1696, 2002. 4. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37: 911–917, 1959. 5. Chauvet C, Bois-Joyeux B, Danan JL. Retinoic acid receptor-related orphan receptor (ROR) alpha4 is the predominant isoform of the nuclear receptor RORalpha in the liver and is upregulated by hypoxia in HepG2 human hepatoma cells. Biochem J 364: 449 –456, 2002. 6. Cho H, Zhao X, Hatori M, Yu RT, Barish GD, Lam MT, Chong LW, DiTacchio L, Atkins AR, Glass CK, Liddle C, Auwerx J, Downes M, Panda S, Evans RM. Regulation of circadian behaviour and metabolism by REV-ERB-a and REV-ERB-b. Nature 485: 123–127, 2012. 7. Chopra AR, Louet JF, Saha P, An J, DeMayo F, Xu J, York B, Karpen S, Finegold M, Moore D, Chan L, Newgard CB, O’Malley BW. Absence of the SRC-2 coactivator results in a glycogenopathy resembling von Gierke’s disease. Science 322: 1395–1399, 2008. 8. Claussnitzer M, Dankel SN, Klocke B, Grallert H, Glunk V, Berulava T, Lee H, Oskolkov N, Fadista J, Ehlers K, Wahl S, Hoffmann C, Qian K, Rönn T, Riess H, Müller-Nurasyid M, Bretschneider N, Schroeder T, Skurk T, Horsthemke B; DIAGRAMⴙConsortium, Spieler D, Klingenspor M, Seifert M, Kern MJ, Mejhert N, Dahlman I, Hansson O, Hauck SM, Blüher M, Arner P, Groop L, Illig T, Suhre K, Hsu YH, Mellgren G, Hauner H, Laumen H. Leveraging cross-species transcription factor binding site patterns: from diabetes risk loci to disease mechanisms. Cell 156: 343–358, 2014. 9. Delezie J, Dumont S, Dardente H, Oudart H, Gréchez-Cassiau A, Klosen P, Teboul M, Delaunay F, Pévet P, Challet E. The nuclear receptor REV-ERBalpha is required for the daily balance of carbohydrate and lipid metabolism. FASEB J 26: 3321–3335, 2012. 10. Dogterom P. Development of a simple incubation system for metabolism studies with precision-cut liver slices. Drug Metab Dispos 21: 699 –704, 1993. 11. Duez H, van der Veen JN, Duhem C, Pourcet B, Touvier T, Fontaine C, Derudas B, Bauge E, Havinga R, Bloks VW, Wolters H, van der Sluijs FH, Vennstrom B, Kuipers F, Staels B. Regulation of bile acid synthesis by the nuclear receptor Rev-erbalpha. Gastroenterology 135: 689 –698, 2008. 12. Duez H, Duhem C, Laitinen S, Patole PS, Abdelkarim M, Bois-Joyeux B, Danan JL, Staels B. Inhibition of adipocyte differentiation by RORalpha. FEBS Lett 583: 2031–2036, 2009. 13. Dussault I, Fawcett D, Matthyssen A, Bader JA, Giguère V. Orphan nuclear receptor RORalpha-deficient mice display the cerebellar defects of staggerer. Mech Dev 70: 147–153, 1998. 14. Everett LJ, Lazar MA. Nuclear receptor Rev-erb␣: up, down, and all around. Trends Endocrinol Metab 25: 586 –592, 2014. 15. Feng D, Liu T, Sun Z, Bugge A, Muliccan SE, Alenghat T, Liu XS, Lazar MA. A circadian rhythm orchestrated by histone deacetylase 3 controls hepatic lipid metabolism. Science 331: 1315–1319, 2011. 16. Grant D, Yin L, Collins JL, Parks DJ, Orband-Miller LA, Wisely GB, Joshi S, Lazar MA, Willson TM, Zuercher WJ. GSK4112, a small molecule chemical probe for the cell biology of the nuclear heme receptor Rev-erb␣. ACS Chem Biol 5: 925–932, 2010. 17. Hamilton BA, Frankel WN, Kerrebrock AW, Hawkins TL, FitzHugh W, Kusumi K, Russell LB, Mueller KL, van Berkel V, Birren BW, Kruglyak L, Lander ES. Disruption of the nuclear hormone receptor RORalpha in staggerer mice. Nature 379: 736 –739, 1996. 18. Jetten AM. Retinoid-related orphan receptors (RORs): critical roles in development, immunity, circadian rhythm, and cellular metabolism. Nucl Recept Signal 7: e003, 2009. 19. Kang HS, Angers M, Beak JY, Wu X, Gimble JM, Wada T, Xie W, Collins JB, Grissom SF, Jetten AM. Gene expression profiling reveals a regulatory role for ROR␣ and ROR␥ in phase I and phase II metabolism. Physiol Genomics 31: 281–294, 2007. 20. Kang HS, Okamoto K, Takeda Y, Beak JY, Gerrish K, Bortner CD, DeGraff LM, Wada T, Xie W, Jetten AM. Transcriptional profiling reveals a role for ROR␣ in regulating gene expression in obesity-associated inflammation and hepatic steatosis. Physiol Genomics 43: 818 –828, 2011. 21. Kim J, Saidel GM, Kalhan SC. A computational model of adipose tissue metabolism: evidence for intracellular compartmentation and differential activation of lipases. J Theor Biol 251: 523–540, 2008.

E113

22. Kohsaka A, Laposky AD, Ramsey KM, Estrada C, Joshu C, Kobayashi Y, Turek FW, Bass J. High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metab 6: 414 –421, 2007. 23. Kojetin DJ, Burris TP. REV-ERB and ROR nuclear receptors as drug targets. Nat Rev Drug Discov 13: 197–216, 2014. 24. Kumar N, Kojetin DJ, Solt LA, Kumar KG, Nuhant P, Duckett DR, Cameron MD, Butler AA, Roush WR, Griffin PR, Burris TP. Identification of SR3335 (ML-176): a synthetic RORalpha selective inverse agonist. ACS Chem Biol 6: 218 –222, 2011. 25. Lau P, Nixon SJ, Parton RG, Muscat GE. RORalpha regulates the expression of genes involved in lipid homeostasis in skeletal muscle cells: caveolin-3 and CPT-1 are direct targets of ROR. J Biol Chem 279: 36828 –36840, 2004. 26. Lau P, Fitzsimmons RL, Raichur S, Wang SC, Lechtken A, Muscat GE. The orphan nuclear receptor, RORalpha, regulates gene expression that controls lipid metabolism: staggerer (sg/sg) mice are resistant to diet-induced obesity. J Biol Chem 283: 18411–18421, 2008. 27. Lau P, Fitzsimmons RL, Pearen MA, Watt MJ, Muscat GE. Homozygous staggerer (sg/sg) mice display improved insulin sensitivity and enhanced glucose uptake in skeletal muscle. Diabetologia 54: 1169 –1180, 2011. 28. Lau P, Tuong ZK, Wang SC, Fitzsimmons RL, Goode JM, Thomas GP, Cowin GJ, Pearen MA, Mardon K, Stow JL, Muscat GE. ROR␣ deficiency and decreased adiposity are associated with induction of thermogenic gene expression in subcutaneous white adipose and brown adipose tissue. Am J Physiol Endocrinol Metab 308: E159 –E171, 2015. 29. Le Martelot G, Claudel T, Gatfield D, Schaad O, Kornmann B, Lo Sasso G, Moschetta A, Schibler U. REV-ERBalpha participates in circadian SREBP signaling and bile acid homeostasis. PLoS One 7: e1000181, 2009. 30. Leroyer S, Tordjman J, Chauvet G, Quette J, Chapron C, Forest C, Antoine B. Rosiglitazone controls fatty acid cycling in human adipose tissue by means of glyceroneogenesis and glycerol phosphorylation. J Biol Chem 281: 13141–13149, 2006. 31. Leroyer S, Vatier C, Kadiri S, Quette J, Chapron C, Capeau J, Antoine B. Glyceroneogenesis is inhibited through HIV protease inhibitor-induced inflammation in human subcutaneous but not visceral adipose tissue. J Lipid Res 52: 207–220, 2011. 32. Mamontova A, Séguret-Macé S, Esposito B, Chaniale C, Bouly M, Delhaye-Bouchaud N, Luc G, Staels B, Duverger N, Mariani J, Tedgui A. Severe atherosclerosis and hypoalphalipoproteinemia in the staggerer mouse, a mutant of the nuclear receptor RORalpha. Circulation 98: 2738 –2743, 1998. 33. Martins-Santos ME, Chaves VE, Frasson D, Boschini RP, Garófalo MA, Kettelhut Ido C, Migliorini RH. Glyceroneogenesis and the supply of glycerol-3-phosphate for glyceride-glycerol synthesis in liver slices of fasted and diabetic rats. Am J Physiol Endocrinol Metab 293: E1352– E1357, 2007. 34. Nye CK, Hanson RW, Kalhan SC. Glyceroneogenesis is the dominant pathway for TG synthesis in vivo in the rat. J Biol Chem 283: 27565– 27574, 2008. 35. Ohoka N, Kato S, Takahashi Y, Hayashi H, Sato R. The orphan nuclear receptor RORalpha restrains adipocyte differentiation through a reduction of C/EBPbeta activity and perilipin gene expression. Mol Endocrinol 23: 759 –771, 2009. 36. Olswang Y, Cohen H, Papo O, Cassuto H, Croniger CM, Hakimi P, Tilghman SM, Hanson RW, Reshef L. A mutation in the peroxisome proliferator-activated receptor gamma-binding site in the gene for the cytosolic form of phosphoenolpyruvate carboxykinase reduces adipose tissue size and fat content in mice. Proc Natl Acad Sci USA 99: 625–630, 2002. 37. Pearen MA, Muscat GE. Orphan nuclear receptors and the regulation of nutrient metabolism: understanding obesity. Physiology 27: 156 –166, 2012. 38. Preitner N, Damiola F, Lopez-Molina L, Zakany J, Duboule D, Albrecht U, Schibler U. The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110: 251–260, 2002. 39. Raspé E, Duez H, Gervois P, Fiévet C, Fruchart JC, Besnard S, Mariani J, Tedgui A, Staels B. Transcriptional regulation of apolipoprotein C-III gene expression by the orphan nuclear receptor RORalpha. J Biol Chem 276: 2865–2871, 2001.

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org

E114

ROR␣ REGULATES ADIPOSE TISSUE GLYCERONEOGENESIS

40. Reshef L, Olswang Y, Cassuto H, Blum B, Croniger CM, Kalhan SC, Tilghman SM, Hanson RW. Glyceroneogenesis and the triglyceride/fatty acid cycle. J Biol Chem 278: 30413–30416, 2003. 41. Shostak A, Meyer-Kovac J, Oster H. Circadian regulation of lipid mobilization in white adipose tissues. Diabetes 62: 2195–2203, 2013. 42. Steinmayr M, André E, Conquet F, Rondi-Reig L, Delhaye-Bouchaud N, Auclair N, Daniel H, Crépel F, Mariani J, Sotelo C, Becker-André M. Staggerer phenotype in retinoid-related orphan receptor ␣-deficient mice. Proc Natl Acad Sci USA 95: 3960 –3965, 1998. 43. Takeda Y, Jothi R, Birault V, Jetten AM. RORgamma directly regulates the circadian expression of clock genes and downstream targets in vivo. Nucleic Acid Res 40: 8519 –8535, 2012. 44. Tordjman J, Chauvet G, Quette J, Beale EG, Forest C, Antoine B. Thiazolidinediones block fatty acid release by inducing glyceroneogenesis in fat cells. J Biol Chem 278: 18785–18790, 2003. 45. Tordjman J, Leroyer S, Chauvet G, Quette J, Tomkiewicz C, Aggerbeck M, Barouki R, Forest C, Antoine B. Cytosolic aspartate aminotransferase, a new partner in adipocyte glyceroneogenesis and an atypical target of thiazolidinedione. J Biol Chem 282: 23591–23602, 2007.

46. Wang Y, Kumar N, Nuhant P, Cameron MD, Istrate MA, Roush WR, Griffin PR, Burris TP. Identification of SR1078, a synthetic agonist for the orphan nuclear receptors RORalpha and gamma. ACS Chem Biol 5: 1029 –1034, 2010. 47. Wang Y, Kumar N, Solt LA, Richardson TI, Helvering LM, Crumbley C, Garcia-Ordonez RD, Stayrook KR, Zhang X, Novick S, Chalmers MJ, Griffin PR, Burris TP. Modulation of retinoic acid receptorrelated orphan receptor alpha and gamma activity by 7-oxygenated sterol ligands. J Biol Chem 285: 5013–5025, 2010. 48. Yang J, Reshef L, Cassuto H, Aleman G, Hanson RW. Aspects of the control of phosphoenolpyruvate carboxykinase. J Biol Chem 284: 27025– 27029, 2009. 49. Yang X, Downes M, Yu RT, Bookout AL, He W, Straume M, Mangelsdorf DJ, Evans RM. Nuclear receptor expression links the circadian clock to metabolism. Cell 126: 801–810, 2006. 50. Yin L, Wu N, Curtin JC, Qatanani M, Szwergold NR, Reid RA, Waitt GM, Parks DJ, Pearce KH, Wisely GB, Lazar MA. Rev-erbalpha, a heme sensor that coordinates metabolic and circadian pathways. Science 318: 1786 –1789, 2007.

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00518.2014 • www.ajpendo.org