Bone Regulates Browning and Energy MetabolismThrough Mature Osteoblast/Osteocyte PPARgExpressionJulia Brun,1 Flavien Berthou,2 Mirko Trajkovski,2 Pierre Maechler,2 Michanlegelo Foti,2 and Nicolas Bonnet1

Diabetes 2017;66:2541–2554 | https://doi.org/10.2337/db17-0116

Peroxisome proliferator–activated receptor g (PPARg) isa master regulator of energy metabolism. In bone, it isknown to regulate osteoblast differentiation and osteo-clast activity. Whether PPARg expression in bone cells,particularly osteocytes, regulates energy metabolism re-mains unknown. Here, we show that mature osteoblast/osteocyte-specific ablation of PPARg in mice (Ocy-PPARg2/2) alters body composition with age, namely, toproduce less fat and more lean mass, and enhances in-sulin sensitivity and energy expenditure compared withwild-type mice. In addition, Ocy-PPARg2/2 mice exhibitmore bone density, structure, and strength by uncouplingbone formation from resorption. When challenged with ahigh-fat diet,Ocy-PPARg2/2 mice retain glycemic control,with increased browning of the adipose tissue, decreasedgluconeogenesis, and less hepatic steatosis. Moreover,thesemetabolic effects, particularly an increase in fatty acidoxidation, cannot be explained by decarboxylated osteocal-cin changes, suggesting existence of other osteokines thatare under the control of PPARg. We further identify bonemorphogenetic protein 7 as one of them. Hence, osteo-cytes coregulate bone and glucose homeostasis througha PPARg regulatory pathway, and its inhibition could beclinically relevant for the prevention of glucose metabolicdisorders.

Bone is a dynamic tissue that undergoes constant remodel-ing, balancing osteoblastic bone formation and osteoclas-tic bone resorption. The transcription factor, peroxisomeproliferator–activated receptor (PPAR)g, stimulates boneresorption through an increase in RANK signaling, leadingto increased c-fos expression and NFATc1 signaling, which

are the essential mediators of osteoclastogenesis and oste-oclast activity (1). In contrast, expression and activation ofPPARg in mesenchymal stem cells promote the differenti-ation of these cells into adipocytes at the expense of theosteoblast lineage (2). Deletion of PPARg in early osteo-blasts, using the early mesenchymal-expressing Col 3.6 kbpromoter, or osterix Cre, increases bone formation and os-teoblast activity (3,4). However, the specific role of PPARgin mature osteoblasts and osteocytes has not yet beenexplored. More recently, bone has been shown to regu-late whole-body glucose homeostasis. Bone is known asan insulin-regulated tissue (5); but more importantly, theexistence of skeletal feedback has emerged with osteocalcin,an osteoblast-derived polypeptide hormone reported to in-crease insulin release from b-cells and indirectly increase in-sulin action through enhanced release of adiponectin fromadipose tissue (6–9). In addition, recent evidence also sug-gests a regulation of glucose metabolism by bone throughosteocalcin-independent mechanisms (10,11). Browning ofthe adipose tissue has received considerable attention toprevent obesity and related comorbidities. Its regulationthrough the PPARs, and particularly, PPARg has been putforward not only in adipose tissue (12,13) but also in liverand skeletal muscle, where it controls the production of sev-eral hepatokines (14) and myokines, respectively (15,16).The fact that bone remodeling occurs daily, in multiplelocations and over extended surfaces, suggests that bonecells have a high energy demand that needs to be met. Re-cently, it has come to light that the skeleton is the fourth-largest glucose-consuming organ after fat, liver, and skeletalmuscle (17). This glucose consumption has always been seento be driven by osteoblast (18) and osteoclast (19) activity.

1Division of Bone Diseases, Department of Internal Medicine Specialties, Facultyof Medicine, Geneva University Hospital, Geneva, Switzerland2Department of Cell Physiology and Metabolism, University of Geneva, and Fac-ulty of Medicine, Centre Médical Universitaire, Geneva, Switzerland

Corresponding author: Nicolas Bonnet, nicolas.bonnet@unige.ch.

Received 30 January 2017 and accepted 29 June 2017.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db17-0116/-/DC1.

© 2017 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, and thework is not altered. More information is available at http://www.diabetesjournals.org/content/license.

Diabetes Volume 66, October 2017 2541

METABOLISM

However, these two bone cells are under the control ofosteocytes, which orchestrate bone modeling/remodelingand calcium homeostasis (20). Moreover, it is the mostabundant cell in the bone tissue. Considering the role ofPPARg in the regulation of energy metabolism and boneremodeling and the high energy expenditure of the skele-ton, we hypothesized that osteocytes might also be key cellsfor regulating energy metabolism. In the current study,we explore the mechanisms by which PPARg expression inmature osteoblasts/osteocytes (OB-OCY) coregulates boneand glucose homeostasis. We demonstrated that osteocyte-targeted deletion of PPARg in mice (Ocy-PPARg2/2) throughthe Dmp1-cre promoter improved glucose homeostasisby increasing browning of adipose tissue and inhibitingthe development of hepatic steatosis, whereas bone frailtywas partially prevented in terms of bone strength but notbone structure, independently of osteocalcin.

RESEARCH DESIGN AND METHODS

Construction of the PPARg floxed allele (PPARgL2) has pre-viously been described (1), and the PPARg-deficient mice,specifically in late OB-OCY, result from a 10-Kb Dmp1-Crerecombination (Ocy-PPARg2/2), which has been crossedonto a full C57BL6J background (21). Control mice (Ocy-PPARg+/+) were constituted by a mix of PPARgLoxP+/+ micewithout DMP1-Cre and PPARgLoxP2/2 mice with DMP1-Cre. The two genotypes of control mice did not exhibitsignificantly different phenotypes. Mice were maintainedon a 12-h light/dark cycle at an ambient temperature of22–25°C. In an aging and high-fat experiment to measuredynamic indices of bone formation, mice were injected withcalcein intraperitoneally at 10 mg/kg at 9 and 2 days beforeeuthanasia. The animals were killed and blood was collectedfor serum measurements.

Aging ExperimentFrom 3 to 6 months of age, Ocy-PPARg+/+ and Ocy-PPARg2/2 male mice were analyzed (n = 8 per group).Six supplemental mice per genotype were used to investi-gate insulin signaling in organs. Mice were injected intra-peritoneally with 150 mU/g insulin (n = 3 per genotype)(NovoRapid; Novo Nordisk) or PBS (n = 3 per genotype)40 min before sacrifice as previously validated (14).

High-Fat ExperimentFrom 4 to 6 months of age, Ocy-PPARg+/+ and Ocy-PPARg2/2 male mice received a high-fat diet (HFD) (60%fat; Kliba 2127 [Supplementary Table 1]) (n = 8 per group).An additional group received a chow diet (10% fat; Kliba2125 [Supplementary Table 1]) for histological analysis (n =6 per genotype).

In Vitro Experiment

Primary Osteoblast CulturesOsteoblasts were isolated from long bones of Ocy-PPARg2/2 and Ocy-PPARg+/+ mice as previously described(22). Osteoblasts were seeded at 60,000 cells per well in

12-well plates. At day 21, RNA was extracted to confirmthe expression of DMP1 and the existence of osteocyticmarkers. Conditioned medium (CM) was collected every24 h.

Primary Adipocyte CultureThe stromal vascular fraction cells from white fat tissueand brown fat tissue were isolated as previously described(23). Primary stromal vascular fraction cells were seeded at300,000 cells/well in 24-well plates. On confluence, the cellswere induced to differentiate for 3 days with DMEM con-taining 10% FCS, 1 mmol/L dexamethasone, 500 mmol/Lisobutylmethylxanthine, 1 mg/mL insulin, 125 mmol/L in-domethacin, 1 nmol/L triiodothyronine (Gibco Life Tech-nology), 1 mmol/L rosiglitazone (Santa Cruz Biotechnology),and CM of OB-OCY culture. Then the induction mediumwas replaced with DMEM containing 10% FCS, 1 mg/mLinsulin, and 1 nmol/L triiodothyronine at days 3 and 5.Cells were stopped at 7 days for Oil Red O and gene expres-sion analyses.

3t3L1 Adipocyte Cell Line Culture3t3L1 adipocyte cell line was cultured as previously de-scribed (24). Cells were seeded at 5,000 cells/well in 24-wellplates. Cells were stopped at 7 days for Oil Red O and geneexpression analyses. At day 0, i.e., at confluency, we splitthe well into different groups: 1) one still under DMEM, 2)culture with Ocy-Pparg+/+ CM, 3) Ocy-Pparg+/+ CM plusrecombinant BMP7 (50 ng/mL), 4) Ocy-Pparg+/+ CM plusrecombinant BMP7 (150 ng/mL), 5) Ocy-Pparg2/2 CM,and 6) Ocy-Pparg2/2 CM plus BMP7-neutralizing antibody(5 mg/mL).

Primary Hepatocyte CulturePrimary hepatocytes were isolated from wild-type (WT)mouse liver as previously described (25). Primary hepato-cytes were seeded at 300,000 cells per well in collagen-coated 12-well plates and cultured in Williams’ Medium Esupplemented with 10% FBS, 1% L-glutamine, 1% penicillin/streptomycin, 1029 mol/L insulin, and 1026 mol/L dexa-methasone. Five hours after plating, cells were incubatedfor 2 days with conditioned media from Ocy-PPARg+/+, Ocy-PPARg2/2, and Ocy-Pparg2/2 plus BMP7-neutralizing an-tibody (5 mg/mL) to OB-OCY. Hepatosteatosis was inducedby adding into these media a mix of 200 mmol/L oleic acid:palmitic acid (ratio 3:1).

Primary IsletPancreatic islets were isolated from male C57BL/6 mice bycollagenase digestion as previously described (26) and cul-tured in RPMI-1640 medium supplemented with 5% heat-inactivated FCS, 10 mmol HEPES, 1 mmol sodium pyruvate,50 mmol 2-mercaptoethanol, and antibiotics. Batches of10 handpicked islets/well were cocultured for 4 days withOcy-PPARg2/2 and Ocy-PPARg+/+ primary OB-OCY. Insulinsecretion measurements were measured by radioimmuno-assay (Linco Research, Inc., St. Charles, MO) using rat in-sulin as standard.

2542 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

InvestigationsFor glucose, insulin, and pyruvate tolerance tests (GTT, ITT,and PTT, respectively), after 6 h starvation, mice wereadministered intraperitoneally with glucose (1.5 g/kg;Grosse Apotheke Dr. G. Bichsel, Interlaken, Switzerland),insulin (0.5 units/kg), or pyruvate (2 g/kg; Sigma), andglycemia was measured from tail blood before the injectionand at 15, 30, 60, 90, and 120 min after the injection. Areaunder the curve (AUC) was calculated relative to baselineglucose for each genotype. Microcomputed tomographyscans (UCT40; Scanco Medical AG, Basserdorf, Switzerland),hyperinsulinemic-euglycemic clamp, 2-[14C]deoxyglucoseuptake, metabolic cage (LabMaster calorimetry; TSE Sys-tems, Homburg, Germany), positron emission tomography(PET)/computed tomography (CT) (Triumph microPET/SPECT/CT system; Trifoil, Chatsworth, CA), histomor-phometry (RM2155 Rotary Microtome; Leica, Wetzlar,Germany), immunohistochemistry, real-time PCR, West-ern blotting, and a three-point bending test (Instron1114; Instron Corp., High Wycombe, U.K.) were per-formed as previously described (14,27,28).

Data AnalysisNormal distribution of the data were verified by the Levenetest. We then tested the effects of genotype in chow dietand HFD by a one-way ANOVA.

To compare the effect of genotype/diet, we used a two-wayANOVA. As appropriate, post hoc testing was performedusing Fisher protected least squares difference. ANCOVAanalyses were performed for the calorimetric data. Differenceswere considered significant at P , 0.05. Data are presentedas mean 6 SEM.

RESULTS

PPARg Deletion in Osteocytes Induced a High–BoneMass and Low–Fat Mass PhenotypePPARg expression increased with osteoblast differentia-tion into osteocytes in vitro and its expression was induc-ible by rosiglitazone (Fig. 1A and B). Immunostainingconfirmed that PPARg was also expressed in osteocytes invivo (Fig. 1C). To achieve a genetic deletion of PPARg in lateosteoblasts/early osteocytes, we crossed Dmp1-Cre mice withhomozygous PPARg-floxed mice to remove the loxP-flankedPPARg alleles only in these cells. Immunohistochemistry,Western blot, and RT-PCR confirmed tissue and cell speci-ficity of PPARg deletion in osteocytes (Fig. 1C–F). As re-ported in the literature, Dmp1-Cre also induced a partialdeletion of PPARg in the skeletal muscle but not in thebrain or intestine. Lengths of femurs or tibiae did not differbetween genotypes (data not shown). However, an exami-nation of the body composition via DEXA for small animals(PIXImus) revealed more femoral bone mineral density(BMD) at the age of 3 months in Ocy-PPARg2/2 comparedwith WT littermates (Ocy-PPARg+/+), which remained sig-nificant with aging (Fig. 2A). With age (6 months), both tra-becular and cortical bone structure respectively at distal andmidshaft femur significantly increased in Ocy-PPARg2/2

versus Ocy-PPARg+/+ mice (Fig. 2B–E). These characteristicsresulted from a higher bone formation rate (BFR) at thetrabecular and periosteal surfaces in Ocy-PPARg2/2 mice,whereas BFR was unchanged at the endosteal surface, andcirculating levels of total and decarboxylated osteocalcinwere similar in Ocy-PPARg2/2 and Ocy-PPARg+/+ mice (Fig.2F–J; Supplementary Table 2). Bone formation effects weremainly attributed to an increase in osteoblast activity ratherthan an increase in osteoblast number, as shown by an in-crease in mineralization apposition rate and an unaffectedmineralization surface on bone surface (MS/BS), mineraliza-tion perimeter on bone perimeter (MPm/BPm), and osteoblastnumbers (Supplementary Table 2). Bone resorption marker,such as carboxyterminal collagen crosslinks (CTX), was sig-nificantly decreased in Ocy-PPARg2/2; regardless, osteo-clast surface on bone surface was not different betweengroups (Fig. 2K and L). For understanding of the mecha-nism by which PPARg deletion in osteocytes increases BFRand decreases resorption, mRNA was extracted from thefemur. Gene expression analyses in the osteocytic fractionof cells showed that Sost (sclerostin) mRNA levels were 33%lower in Ocy-PPARg2/2 compared with Ocy-PPARg+/+ (Fig.2M). Apoptotic gene expression such as Caspase 3, Bcl2,and Bax were unchanged (Supplementary Table 3). In theosteoblastic fraction, both Runx2 and Opg levels were sig-nificantly higher in Ocy-PPARg2/2: 19.2% and 197.6%,respectively, versus Ocy-PPARg+/+ (Fig. 2N and O). Further-more, primary mesenchymal stem cells from Ocy-PPARg2/2

differentiated in vitro into more mature osteoblasts, asshown by higher alkaline phosphatase levels in Ocy-PPARg2/2 after 21 days of culture when Dmp1 wasexpressed (Fig. 2P). Furthermore, gene expression of osteo-blast markers such as Opg/Rankl, Runx2, osteocalcin(Bglap), Glut1, Glut 4, Fox2, PPARa/d, and Bmp7 washigher in Ocy-PPARg2/2 compared with Ocy-PPARg+/+,whereas E11, an early osteocyte marker, was lower (Fig.2Q). These data indicated that osteocyte expression ofPPARg regulates both osteoblast differentiation/functionand osteoclast resorption.

PPARg Deletion in Late OB-OCY Increased EnergyMetabolismAt 3 months of age, lean and fat mass percentages as well asglucose levels measured in response to GTT were unaffectedby PPARg deletion (Fig. 3A–C; Supplementary Fig. 1). How-ever, with age (6 months), Ocy-PPARg2/2 gained less fatmass and more lean mass than Ocy-PPARg+/+ with a similarbody weight gain, respectively, of 12.0% and 12.5% (Fig.3A–C). At sacrifice, organ weights indicated a lower liver andepididymal white adipose tissue (eWAT) in Ocy-PPARg2/2

compared to Ocy-PPARg+/+ (Supplementary Fig. 2). Glycemiccontrol was better in Ocy-PPARg2/2 mice compared withOcy-PPARg+/+, with enhanced glucose tolerance (analyzedby GTT) and insulin sensitivity (measured by ITT) (Fig. 3D–F).Metabolic cage profiling showed more VO2 consumption, ahigher respiratory exchange ratio (RER), and more heat pro-duction in Ocy-PPARg2/2 mice compared to Ocy-PPARg+/+,

diabetes.diabetesjournals.org Brun and Associates 2543

mainly during the night (Fig. 3G and H); these differencesremained significant after division by lean mass or by per-formance of an ANCOVA (Supplementary Fig. 3A–D).Movement and cumulative food intake were unchanged(Supplementary Fig. 4A and B). RER was significantly higherin Ocy-PPARg2/2 compared with Ocy-PPARg+/+ (0.91 60.008 vs. 0.86 6 0.01, P , 0.01), with values indicating amix of fat and carbohydrate as fuel sources. In order tofurther investigate the higher energy expenditure of Ocy-PPARg2/2, we evaluated the body temperature of differentregions of interest. Internal body temperature, i.e., eye tem-perature measurement, and temperature in the musculo-skeletal limb were higher in Ocy-PPARg2/2 comparedwith Ocy-PPARg+/+ mice (Fig. 3I). Analyses of the ventraland dorsal infrared images showed that the interscapularregions, i.e., brown adipose tissue (BAT), contributed to the

overall higher temperature in Ocy-PPARg2/2 comparedwith inguinal fat, i.e., white adipose tissue (WAT) (Fig. 3I).However, body temperature response to 24 h cold exposurewas similar between genotypes (Supplementary Fig. 4C). Tofurther address the metabolic effects of PPARg deletion inOB-OCY, we performed a hyperinsulinemic-euglycemic clampin conscious and unrestrained mice. Ocy-PPARg2/2 miceexhibited a marked increase in the glucose infusion ratesneeded to maintain the clamped glucose levels, demonstrat-ing an increase of insulin sensitivity (Fig. 3J). In accordance,insulin staining in the pancreas showed smaller islet size inOcy-PPARg2/2 compared with Ocy-PPARg+/+, whereas thenumber remained unchanged (Supplementary Fig. 4D–G).Moreover, phosphorylation level of Aktser473, considered animportant downstream target of insulin signaling, was sig-nificantly higher in bone, skeletal muscle, BAT, and liver in

Figure 1—PPARg expression in IDG-W3 osteocyte cell lines and specific deletion of PPARg in osteocyte using DMP1-cre promotor. A: Timecourse expression of Dmp1 in IDG-W3 osteocyte cell line mark with GFP under Dmp1 promoter. D, day. (n = 3). B: Time courseexpression of PPARg in IDG-W3 osteocyte cells treated or not with rosiglitazone (Rozi) (n = 3). C: Immunohistochemical staining of PPARg incross-sectional tibia at the midshaft (upper panel, bar = 20 mm) and in longitudinal sections of the proximal tibia (lower panel, bar = 40 mm)showing the absence of staining in osteocytes of the cortical and trabecular compartments (*), whereas the staining remains in osteoclasts ($)(n = 4). D: PPARg expression in the bone fractions ofOcy-PPARg2/2 andOcy-PPARg+/+mice (n = 4). E: Western blot of Dmp1-Cre, total PPARg,and Pan actin (n = 3). F: Western blot quantification by ImageJ of PPARg protein expression in Ocy-PPARg2/2 reported to Ocy-PPARg+/+

(n = 3). Bars show means (6SEM). Closed bars, Ocy-PPARg2/2; open bars, Ocy-PPARg+/+. *P < 0.05, **P < 0.01, significant difference vs.Ocy-PPARg+/+.

2544 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

Figure 2—Bone phenotype of PPARg-deficient mice in late OB-OCY at 6 months of age. A: Femoral BMD. B and C: BV/TV and trabecularthickness (Tb.Th) at the distal femur.D and E: Ct.TV and Ct.BV at midshaft femur (n = 8 per group). F: Images illustrate double fluorescent calceinlabels on trabecular surfaces (n = 6 per group; bar = 50 mm). G and H: Bone formation indices at periosteal (Ps) and endocortical (Ec) surfaces.BFR and bone perimeter (BPm) (n = 6 per group). I and J: Total and undercarboxylated osteocalcin (osteocalcin glu) measured by ELISA inserums. K: Osteoclast surface on bone surface (OcS/BS) (n = 8 per group). L: Serum collagen type I crosslinked C-telopeptide (CTX) measuredby ELISA assay in serums. M–O: Relative mRNA gene expression from flushed femur after isolation of osteocyte fraction (M) and osteoblastfraction (N and O) (n = 4 per group). P: Alkaline phosphatase (ALP) production in mesenchymal stem cell after 9, 15, or 21 days of culture inosteogenic condition (n = 4 per group). Q: Relative mRNA gene expression after 21 days of cultures from Ocy-PPARg2/2 and Ocy-PPARg+/+

femurs (n = 4 per group). Bars show means (6SEM). *P < 0.05, **P < 0.01, ***P < 0.001 significant difference vs. Ocy-PPARg+/+.

diabetes.diabetesjournals.org Brun and Associates 2545

Figure 3—At 6 months of age, PPARg deletion in late osteoblast/osteocyte controls glucose homeostasis by increasing metabolic rate andglucose uptake in WAT, BAT, and bone and improves insulin sensitivity. A–C: Body composition analyzed in vivo by PIXImus; body weight (A),fat mass percentage (%) (B), and lean mass percentage (C) (n = 8 per group). D: Glucose AUC obtained during a GTT (n = 8 per group). E: GTT at6 months (n = 8 per group). F: ITT (n = 8 per group). G and H: Energy metabolism investigation performed in metabolic cages during twoconsecutive days and nights (n = 6 per group); VO2 and AUC of RER (G) and heat production (H). I: Body temperature evaluated by infraredcamera. Note the higher temperature of the limb ofOcy-PPARg2/2, indicated by the white spots (n = 8 per group). J and K: Glucose infusion rate(GIR) and tissue-specific 2-[14C]deoxyglucose (2-[14C]DG) in iWAT and BAT during hyperinsulinemic-euglycemic clamp in awake mice (n =6 per group). L: Standardized uptake values (SUVs) of the radiolabeled tracer 18FDG; note the higher 18FDG uptake in bone of the limb ofOcy-PPARg2/2, indicated by yellow spots (n = 4–5 per group; bar = 1 mm). Bars show means (6SEM). Closed bars or continuous lines, Ocy-PPARg2/2; open bars or dashed lines, Ocy-PPARg+/+. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference vs. Ocy-PPARg+/+.

2546 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

Ocy-PPARg2/2 compared with Ocy-PPARg+/+ mice andtended to be in WAT (Supplementary Fig. 4H and I). Toinvestigate peripheral glucose uptake, we coadministered2-[14C]deoxyglucose during the clamp. Whereas no changeswere observed in skeletal muscle, there was an increaseduptake from the inguinal (i)WAT, the tibia, and BAT in Ocy-PPARg2/2 compared with Ocy-PPARg+/+ mice (Fig. 3K).These results were confirmed by microPET, resulting inmore accumulation of fluodeoxyglucose (18F) (18FDG)in long bones of Ocy-PPARg2/2 (Fig. 3L), whereas no sig-nificant differences were observed in quadriceps and gas-trocnemius.

Lack of PPARg in Mature OB-OCY Increased AdiposeBrowningWe next investigated whether the higher consumption ofglucose by adipose tissue originated from a difference in ad-ipocyte density and volume. Measuring adipocyte size distri-bution revealed an increased number of small adipocytes and

a decreased number of large adipocytes in eWAT of Ocy-PPARg2/2 (Fig. 4A and B). Moreover, morphologically,white fat depots had more of a multinodular pheno-type in Ocy-PPARg2/2 compared with Ocy-PPARg+/+ mice(Fig. 4A). These features are characteristic of beige adipo-cyte formation within WAT as well as mitochondrial activityof BAT. Immunohistochemistry confirmed the expressionof uncoupling protein 1 (UCP1) in WAT of Ocy-PPARg2/2

but not in Ocy-PPARg+/+ (Fig. 4A and B). In BAT depots, weobserved dense UCP1 staining in both genotypes; however,larger white droplets accumulated in the BAT depots inOcy-PPARg+/+ (Fig. 4C). Fat depots of the Ocy-PPARg2/2

mice showed increased expression of brown adipocyte–spe-cific markers compared with Ocy-PPARg+/+, particularly ininguinal WAT (Fig. 4D–F). In vitro, primary BAT culturesshowed larger numbers of lipid droplets in Ocy-PPARg2/2,with a smaller size than in Ocy-PPARg+/+ (Fig. 4G). BAT andinguinal WAT primary culture also confirmed darker UCP1staining in Ocy-PPARg2/2 compared with Ocy-PPARg+/+

Figure 4—At 6 months of age, PPARg deletion in mature OB-OCY promotes browning of the WAT and BAT activity. A: Hematoxylin-eosin (H&E)and UCP1 staining on sections from eWAT. Arrows indicate accumulation of small lipid droplets. $More brown staining in Ocy-PPARg2/2 vs. Ocy-PPARg+/+ (n = 6 per group). B: Cell-size profiling of adipocytes from eWAT; points show mean of pooled fractions from eachanimal (n = 6 per group with 4 sections per animals). C: Hematoxylin-eosin and UCP1 staining on sections from BAT (n = 6 per group with3 sections per animals). D–F: Relative mRNA gene expression in eWAT (D), iWAT (E), and BAT (F) (n = 4 per group).G: Primary culture of BAT; OilRed O staining after 6 days of culture indicated larger numbers of small lipid droplets (n = 6 per group; bar = 100 mm). H: Immunohistochemistryon BAT and iWAT, automated quantification of UCP1 (n = 4 per group; bar = 100 mm). Bars show mean (6SEM). OD, optical density. *P< 0.05,**P < 0.01, ***P < 0.001 significant difference vs. Ocy-PPARg+/+.

diabetes.diabetesjournals.org Brun and Associates 2547

(Fig. 4H), thus arguing a browning of WAT and more activ-ity of BAT in Ocy-PPARg2/2 mice.

Ocy-PPARg2/2 Mice Partially Protected Against theDysmetabolism Induced by High Dietary Fat IntakeWe next challenged the Ocy-PPARg2/2 mice to an HFD orcontrol diet (CD) to investigate whether these mice areprotected against diet-induced bone fragility, obesity, andhyperglycemia. In Ocy-PPARg+/+ mice, HFD did not affectendocortical bone formation rate but increased periostealBFR, perhaps as a result of body weight increase (Sup-plementary Fig. 5A and B). Ocy-PPARg2/2 mice on HFDexhibited a higher endocortical bone formation rate than Ocy-PPARg+/+ (Supplementary Fig. 5B). Hence, Ocy-PPARg2/2

HFD mice have more cortical bone volume on tissue volume(Ct. BV/TV) and cortical thickness (Ct.Th) than Ocy-PPARg+/+

mice (Supplementary Fig. 5C–E). After biomechanicalstrength testing using a three-point bending test, weobserved that Ocy-PPARg2/2 mice exhibited more ultimateforce, stiffness, and plastic and fracture energy comparedwith Ocy-PPARg+/+ (Supplementary Table 4). In Ocy-PPARg+/+ mice, HFD increased elastic and decreased plasticenergy, whereas Ocy-PPARg2/2 mice were unaffected byHFD (Supplementary Table 4). An HFD regimen increasedbody weight in both genotypes over time (Fig. 5A). How-ever, food and hydric recording over 24 h indicated moreintake in Ocy-PPARg2/2 compared with Ocy-PPARg+/+

mice (Supplementary Fig. 6A and B). Nevertheless, the finalbody weight observed in Ocy-PPARg2/2 did not reach thevalues of Ocy-PPARg+/+ mice, suggesting that Ocy-PPARg2/2 animals were partially protected from HFD(Fig. 5A). EchoMRI scans after 8 weeks of HFD showedthat fat mass increase was prevented in Ocy-PPARg2/2

(increase of 17% vs. CD and 44% vs. CD in Ocy-PPARg+/+,P , 0.05), whereas lean mass was greater (14.2% vs. CDand 9.6% vs. CD in Ocy-PPARg+/+) (Fig. 5B and C). OnHFD, movement, VO2, and heat were increased in Ocy-PPARg2/2 (41%, 13%, and 13% vs. Ocy-PPARg+/+, P ,0.05) (Fig. 5D–F) compared with the CD experiment; geno-type differences on HFD disappeared after division ofparameters per lean mass or ANCOVA analysis(Supplementary Fig. 6C–F). RER was not significantly dif-ferent between genotypes and mean value decreased com-pared with CD, indicating that fat is the predominant fuelsource (Supplementary Fig. 6G). Body temperature was alsohigher, particularly in the BAT neck region (1.5% vs. Ocy-PPARg+/+, P, 0.01) (Supplementary Fig. 7A). Responses toGTT, ITT, and PTT were improved in Ocy-PPARg2/2

(AUC 222%, 29%, and 213%, respectively, vs. Ocy-PPARg+/+, P , 0.1) (Fig. 5G–I). To further address themetabolic effects of PPARg deletion in OB-OCY, we per-formed a hyperinsulinemic-euglycemic clamp after 8 weeksof the dietary challenge. Ocy-PPARg2/2 mice exhibited thesame trend of a higher glucose infusion rate required tomaintain the clamped glucose levels, demonstrating an in-crease of insulin sensitivity (Fig. 5J). Measurement ofinsulin levels before and after clamping confirmed the

increase in insulin levels during clamping and illustrated nosignificant difference between genotypes (SupplementaryFig. 7B). The difference between glucose disappearance(Rd) before and after the clamp was higher in Ocy-PPARg2/2

mice, suggesting less hepatic glucose production (Supple-mentary Fig. 7C). We coadministered 2-[14C]deoxyglucoseduring the clamp. No changes were observed in glucoseuptake in skeletal muscle and WAT, although there was asignificant increased uptake by the tibia and a trend for theBAT in Ocy-PPARg2/2 compared with Ocy-PPARg+/+ mice(Supplementary Fig. 7D). These results were confirmed bymicroPET, with more accumulation of 18FDG in BAT and inthe long bones of Ocy-PPARg2/2 (Fig. 5K and L).

Lack of PPARg in Late OB-OCY Prevented FatInfiltration in Skeletal Muscle and Liver and IncreasedAdipose BrowningOn CD, histology of the gastrocnemius revealed homoge-nous fiber size distribution, polygonal shape of fibers, andnumber of peripheral nuclei in both genotypes (Fig. 6A). OnHFD, Ocy-PPARg2/2 seemingly prevented fibrosis in thetissue, with less mononucleated inflammatory cells, centralnucleation of the myofibers, and fat infiltration comparedwith Ocy-PPARg+/+ mice. In accordance, gene expressionanalysis in HFD showed more muscle regeneration andfatty acid (FA) oxidation, indicated by high expression ofMef2b and PPARa/d, and muscle force in Ocy-PPARg2/2

(Fig. 6B and C). Liver steatosis, as illustrated by accumula-tion of white lipid droplets stained with Oil Red O, was lesspronounced in Ocy-PPARg2/2 compared with Ocy-PPARg+/+

on HFD (Fig. 6D). Gene expression indicated reducedFA transporters in Ocy-PPARg2/2, in particular Cd36,which was confirmed by Western blot (Fig. 6E and F), sug-gesting a reduced uptake of FA from the bloodstream intothe liver. Moreover, it also indicated less lipogenesis inOcy-PPARg2/2 through the lower expression of Fas, Acc1,and Acc2 and a decrease in gluconeogenesis through a de-crease in Ppeck expression (Fig. 6E). FA oxidation, exportand cholesterol were modestly affected (Supplemen-tary Fig. 7E). Adipocyte morphology and beige adipogenicgene markers including Ucp1, Cox5b, and Dio2 indicated abrowning of the eWAT in Ocy-PPARg2/2 (Fig. 6G). Fur-thermore, in the BAT, Ocy-PPARg2/2 exhibited lesslarge droplet accumulation compared with Ocy-PPARg+/+

on HFD (adipocyte with lipid droplet size 2,000–2,400;5.3 6 0.2% compared with 8.3 6 0.5%, respectively, P ,0.01). Gene expression analysis showed more capabilityof lipid oxidation in Ocy-PPARg2/2 versus Ocy-PPARg+/+

through the increased expression of PPARa/d/g, Elovl3,Dio2, Ucp1, and Lpl (Fig. 6H).

Osteokines Secreted by Ocy-PPARg2/2 Mice ControlAdipose, Liver, and Pancreas FunctionsTo investigate the proof of concept that deletion of PPARgin bone cells is able to prevent hepatic steatosis, improve FAmetabolism by adipocytes, and augment pancreatic islet in-sulin secretion, we applied CM or performed real cocul-tures of mature osteoblasts with hepatocytes, adipocytes,

2548 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

and pancreatic b-cells. For this purpose, primary OB-OCYof Ocy-PPARg2/2 and Ocy-PPARg+/+ were extracted fromlong bones as previously described (22). After 21 days ofculture, CM was collected and added to primary hepatocytesfrom WT mice incubated with a mix of oleate and palmitatefor 48 h to induce steatosis (Fig. 7A). Lipid content in CMof Ocy-PPARg+/+ and Ocy-PPARg2/2 OB-OCY was not de-tectable, arguing a similar exposure of lipid on hepatocytes.The effects on lipid metabolism were analyzed by gene

expression and Oil Red O staining. Lipid droplet accumula-tion was lesser in hepatocytes exposed to Ocy-PPARg2/2

versus Ocy-PPARg+/+ CM (Fig. 7B). Cd36 FA transporterexpression was also lower with Ocy-PPARg2/2 CM(251% vs. no control media and 252% vs. Ocy-PPARg+/+

CM, all P , 0.001), and Ppara gene expression was higher(48% vs. control media and 22% vs. Ocy-PPARg+/+ CM, allP , 0.05) (Fig. 7C). We then tested the effect of CM on3T3L1 adipocyte cell lines. Exposure to Ocy-PPARg2/2 CM

Figure 5—PPARg deletion in late OB-OCY improves glucose homeostasis by increasing metabolic rate and improving insulin sensitivity andglucose uptake by BAT and bone after 8 weeks of HFD. A: Body weight (n = 8 per group). B andC: Fat and lean mass evaluated by echoMRI (n =8 per group); $P < 0.05 significant difference vs. chow diet. From D to L, all the mice are on HFD. D–F: Energy metabolism investigationperformed in metabolic cages during 1 day and night (n = 6 per group); movement (D), VO2 (E), heat production (F). G: GTT (n = 8 per group). H:ITT (n = 8 per group). I: PTT (n = 8 per group). J: Glucose infusion rate (GIR) during hyperinsulinemic-euglycemic clamp in awake mice (n = 6 pergroup). K and L: Standardized uptake values (SUV) of the radiolabeled tracer 18FDG in bone (bar = 1 mm) and BAT (bar = 1 cm); note the higher18FDG uptake in bone and BAT ofOcy-PPARg2/2 indicated by red and white spots (n = 5 per group). Arrows indicate 18FDG uptake in the BAT.Bars show means (6SEM). HF, high-fat. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference vs. Ocy-PPARg+/+.

diabetes.diabetesjournals.org Brun and Associates 2549

improved lipid metabolism as shown by Oil Red O quanti-fication, and the gene expression profile was similar to theone described in vivo with an increase of FA synthesis and

oxidation, respectively, illustrated by an increase in lipopro-tein lipase, Fas, and Ppara/d/g (Fig. 7D). Immunostainingand Western blot analysis indicated an increased expression

Figure 6—PPARg deletion in late OB-OCY prevents fat infiltration in skeletal muscle and steatosis and improves browning and BAT activity after8 weeks of HFD. A: Hematoxylin-eosin staining on sections of gastrocnemius (n = 6 per group with 3 sections per animals; bar = 40 mm).B: Relative mRNA gene expression for HFD in gastrocnemius (n = 4 per group). C: Limb force evaluated by handgrip for HFD groups (n = 8 pergroup). D: Oil Red O staining on sections of liver; note the larger accumulation of lipid droplets inOcy-PPARg+/+ vs. Ocy-PPARg2/2 (n = 6 per group;bar = 50 mm). E: Relative mRNA gene expression of FA transports, FA synthesis, and gluconeogenesis markers in liver on HFD (n = 4 per group).F: Western blot of CD36 and quantification for CD and HFD (n = 3 per group). $$$P < 0.001 significant difference vs. CD. G: UCP1 staining onsections from eWAT and relative mRNA gene expression for HFD (n = 6 per group with 3 sections per animals; bar = 100 mm). H: UCP1 stainingon sections from BAT; arrows indicate less accumulation of big lipid droplets in BAT of Ocy-PPARg2/2 vs. Ocy-PPARg+/+ and relative mRNAgene expression for HFD (n = 6 per group with 3 sections per animals; bar = 100 mm). Bars show means (6SEM). Black bars, Ocy-PPARg2/2;white bars, Ocy-PPARg+/+. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference vs. Ocy-PPARg+/+.

2550 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

of UCP1 in cells exposed to Ocy-PPARg2/2 CM comparedwith WT. Coculture of mature OB-OCY Ocy-PPARg2/2

was also performed with pancreatic islets isolated from Ocy-PPARg+/+ mice, in the presence of stimulatory glucose(28 mmol/L) for 6 h. Insulin release was quantified byradioimmunoassay. Coculture with OB-OCY Ocy-PPARg2/2

increased insulin secretion by pancreatic islets (130% vs. noOB and 83% vs. OB-OCY Ocy-PPARg+/+, both P , 0.01)(Fig. 7G and H). In order to confirm that these effects wereindependent of osteocalcin, as shown in vivo, we evaluatedtotal and decarboxylated osteocalcin in CM of OB-OCY Ocy-PPARg2/2 and Ocy-PPARg+/+; no differences were observed(Fig. 7I). Based on literature and gene expression profiles(Fig. 2Q), we analyzed BMP7 levels in the CM and serum ofthe aging experiment by ELISA assay and observed higherlevels in Ocy-Pparg2/2 (respectively, 8.4% and 107% vs.Ocy-Pparg+/+, P , 0.05) (Fig. 8A and B). In turn, the

addition of BMP7-neutralizing antibody to Ocy-Pparg2/2

CM blocked the steatosis prevention in hepatocytes (Fig.8C and D) and FA oxidation in 3t3L1 adipocytes (Fig. 8Eand F). Moreover, when we added recombinant BMP7 toOcy-Pparg+/+ CM, we partially simulated the effect of Ocy-Pparg2/2 CM with an increase in Oil Red O staining andPpard/g gene expression (Fig. 8G and H). Finally, in vivoadministration of a BMP7-neutralizing antibody blockedthe hypoglycemia described in Ocy-Pparg2/2 mice (Fig. 8I).

DISCUSSION

PPARg is a master transcriptional factor that is expressedin many tissues, playing both common and specific roleswherever it is expressed. Here we show that PPARg in boneand more specifically in mature OB-OCY acts as a mediatorof systemic energy metabolism and coregulates bone andglucose homeostasis. PPARg deletion results in an increase

Figure 7—OB-OCY secrete a factor regulating steatosis, lipid oxidation, and insulin secretion. A: Schematic of the experiment performed onprimary hepatocyte. B: Images illustrating less lipid droplets in hepatocytes exposed to Ocy-PPARg2/2 OB-OCY CM (n = 3 per group, duplicateexperiment). Bar = 50 mM. C: Relative mRNA gene expression of CD36 transporter (C1) and PPARa (C2) (n = 4 per group, duplicate experiment).D–F: Effects of 6 days of Ocy-PPARg2/2 CM on 3T3L1 cell line differentiation in adipocytes (n = 6 per group, duplicate experiment). Oil Red Oquantification (D1), relative mRNA gene expression (n = 4 per group, duplicate experiment) (D2), UCP1immunostaining (n = 4 per group, bar =100 mm) (E), and Western blot of UCP1 (n = 4 per condition) (F). G: Schematic of the coculture performed between late OB-OCY and isolatedislets (n = 8–10 per group, duplicate experiment). H: Insulin released from cocultured islets stimulated with 28 mmol/L glucose (n = 8–13 pergroup). I: Total and undercarboxylated osteocalcin (osteocalcin glu) measured by ELISA in CM (n = 10 per group). Bars show means (6SEM).Black bars, Ocy-PPARg2/2; white bars, CM of the original cell (hepatocyte, adipocyte, or islet); dark-gray bars, Ocy-PPARg+/+. OA/PA, oleicacid/palmitic acid. OD, optical density. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference.

diabetes.diabetesjournals.org Brun and Associates 2551

in bone formation and a decrease in bone resorption duringbone mass acquisition, thereby increasing BMD in young-adult rodents. In accordance with the pharmacological workof Marciano et al. (29), repression of PPARg promotesosteogenesis in part through the increased production andsecretion of members of the bone morphogenetic proteinfamily. These effects are opposite those of PPARg activa-tors, such as thiazolidinediones, which inhibit bone forma-tion and promote osteocyte apoptosis through an increasein sclerostin expression, a Wnt antagonist (30,31). Despitethe increase in bone formation index, both total and decar-boxylated osteocalcin were unchanged in the serums. Inhi-bition of PPARg decreased the bone resorption marker,CTX, through an increased Opg/RankL ratio, supportingthe notion that PPARg inhibition in osteocytes is able toorchestrate an uncoupling effect in bone. Interestingly, thecomparative increase in BMD, trabecular and cortical micro-architecture, and bone strength in Ocy-PPARg2/2 mice ac-quired from an early age remained throughout aging. Ourstudy meaningfully highlights the important role played byPPARg in Dmp1-expressing cells, predominantly osteocytes,contrasting with the absence of a drastic phenotype in thePPARg-flox col3.6-cre mice (3). In addition to the bone phe-notype, mice deficient for PPARg in Dmp-1–expressing cellshad improved body composition at 6 months of age, gainedless fat, and lost less lean mass. They exhibited a high RERand metabolic rate independently of lean mass. Strikingly,

body temperature was higher, particularly in the BAT-richneck region and musculoskeletal limb region, showing thatdeletion of PPARg in bone improved thermogenic capacitywithout changing tbone response to acute cold exposure.In addition, WAT was clearly being converted to BAT(known as beige), a mechanism well known to reduce obe-sity and hyperglycemia (28). 2-[14C]deoxyglucose and18FDG distributions indicated more glucose consumptionin the bone and BAT of Ocy-PPARg2/2 mice as a result ofan increase in insulin sensitivity. Given that the differencein glucose levels during the GTT was greater than thatobserved for ITT or clamps in vitro lets us suggest thatOcy-PPARg2/2 mice may have enhanced glucose-stimulatedinsulin secretion. Unfortunately, we were unable to test thisdirectly, owing to insufficient blood samples; therefore, thisshould be tested in future studies in order to solve thediscrepancies with the lesser insulin content seen in islets.Hence, deletion of PPARg in bone improved glucose homeo-stasis by targeting mainly fat and liver metabolism withoutany major effect on muscle. The absence of a metabolicphenotype at 3 months of age emphasizes our previousfinding of the important role of PPAR in the aging processwith an exponential increase around the age of 4 months(32). When subjected to an HFD, Ocy-PPARg2/2 micegained less weight than WT littermates, and the AUC ofOcy-PPARg2/2 challenged with an HFD was equivalent tothat of WT mice on the CD (normal chow). In these

Figure 8—OB-OCY secrete BMP7, regulating steatosis and lipid oxidation, respectively, in hepatocyte and adipocyte. A and B: BMP7measuredby ELISA assay in CM and serum of Ocy-PPARg2/2 and Ocy-PPARg+/+ mice aged 6 months. C and D: Images and quantification of lipiddroplets in hepatocytes exposed to Ocy-PPARg+/+ CM, Ocy-PPARg2/2 CM, and Ocy-PPARg2/2 CM with or without BMP7-neutralizingantibody (Ab-BMP7) (n = 3 per group, duplicate experiment). Nb, number. E and F: Oil Red O quantification in 3T3L1 cell lines after 6 daysof exposure to Ocy-PPARg+/+, Ocy-PPARg2/2, and Ocy-PPARg2/2 CM 6 Ab-BMP7. F: Relative mRNA gene expression. G and H: Oil Red Oquantification in 3T3L1 cell lines after 6 days of exposure to Ocy-PPARg+/+ CM with or without two doses of recombinant BMP7 (50 and150 ng/mL). H: Relative mRNA gene expression. I: Glucose levels after two intravenous injections of Ab-BMP7 (273 ng/30 g body wt every2 days) as previously described (38). Glucose was evaluated 6 h after the last injections. Bars show means (6SEM). Black bars, Ocy-Pparg2/2;white bars, CM of the original cell (hepatocyte, adipocyte, or islet); dark-gray bars, Ocy-Pparg+/+; light-gray bars, Ocy-Pparg2/2 plusBMP7-neutralizing antibody. OD, optical density. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference.

2552 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017

conditions, Ocy-PPARg2/2 mice exhibited less fragile bone,steatosis, white and adipose mass, and more muscle force andbrown adipose activity compared with WT. In vitro experimentsusing cocultures of bone and other cells confirmed that theconditioned media from primary OB-OCY cultures of Ocy-PPARg2/2 mice have the capacity to decrease hepatocyticsteatosis, increase insulin secretion by b-cells, and increaselipid oxidation by adipocytes. Hence, late OB-OCY secretedsome osteokines regulated by PPARg, which secondarilyaffected hepatocytes, b-cells, and adipocytes. Osteocalcinwas the first osteokine able to control glucose homeostasis(10,11,33–35). However, both total and undercarboxylatedosteocalcin were unchanged both in vivo and in vitro. Inaddition, compared with osteocalcin these osteokines seemedto not affect muscle metabolism. Sclerostin is able to con-trol anabolic metabolism in adipocytes. More specifically,Sost2/2 mice have a reduction in adipocyte size, with animprovement of glucose tolerance and enhancement of in-sulin sensitivity in WAT with concomitant increase ofPPARgc1a and Ucp-1, i.e., markers of browning (36).From our gene expression analysis, BMP7 increase in Ocy-PPARg2/2 mice could also explain this metabolic pheno-type. BMP7 is known to stimulate browning and reducesteatosis through an endocrine mechanism (37–40). In-terestingly, BMP7 levels in the CM and serum of Ocy-Pparg2/2 were also higher compared with Ocy-Pparg+/+.Moreover, adding BMP7-neutralizing antibody in Ocy-Pparg2/2 CM did block its effects on FA oxidation inhepatocytes and adipocytes. Last, administration of a BMP7-neutralizing antibody in vivo blocked the hypoglycemia de-scribed in Ocy-Pparg2/2 mice. Considering the molecularweight of BMPs from 10 to 50 kDa, it is very plausiblethat osteocytes actively secrete BMP through the lacuna-canalicular system, as we now know that small proteins(,70 kDa) circulate not only through the osteocytelacunocanalicular network (41–43) but also secondarilyto other tissues and act as a mediator of systemicmetabolism (44).

There are several limitations to our study. Firstly, weused the Dmp1-cre promotor to delete PPARg. Dmp1-Crehas previously been used as a specific promotor for lateOB-OCY; however, more recently it has also been shownto be slightly expressed in skeletal muscle (45,46). Notwith-standing, Ocy-PPARg2/2 mice did not exhibit major skele-tal muscle phenotype, as indicated by movement data,hyperinsulinemic-euglycemic clamp, and/or PET/CT analy-sis. Moreover, if deletion of PPARg were dominant in skel-etal muscle, which was not the case as indicated by Westernblot and quantitative RT-PCR, we should have insulin re-sistance (47). On the contrary, we had an increase in insulinsensitivity, arguing definitively that the small deletion ofPPARg in muscle using the Dmp1-Cre system does nothave a major effect in our model. Contrary to Lim et al.(48), we did not see any expression of Dmp1-Cre in thebrain or intestine. Secondly, we were not able at this timepoint to elucidate all osteokines regulated by PPARg thatcan explain the energy metabolism of our Ocy-PPARg2/2

mice. However, we know that it is not explained by changesin total or decarboxylated osteocalcin and that BMP7 ispartially involved. Increased expression and/or activationof PPARg in bone tissue has been involved in the patho-physiology of diabetes-induced bone fragility (32). There-fore, our new findings suggest that PPARg in lateOB-OCY not only contributes to regulate bone remodelingbut also plays a critical function in glucose metabolismthrough both an increase of glucose uptake by the bonetissue itself and targeting key organs of energy metabolism,such as pancreas, adipose tissue, and liver, by secretion ofosteokines other than osteocalcin, such as BMP7. Hence,this work brings new insights into the role that the skeletaltissue could play in pathophysiological processes leading toobesity, metabolic syndrome, and diabetes.

Acknowledgments. The authors thank Serge Ferrari (Geneva UniversityHospitals, Switzerland) for discussions and critical reading, Lynda Bonewald(University of Missouri–Kansas City) for providing the Dmp1-cre, and BeatriceDesvergne (University of Lausanne, Lausanne, Switzerland) for the PPARgL2mice. The authors thank Madeleine Lachize, Juliette Cicchini, and Pierre Apos-tolides for technical assistance with the bone tissue investigation (Servicedes Maladies Osseuses, University of Geneva, Switzerland). As a platform,the authors thank Christelle Veyrat-Durebex (Département de Physiologie Cellulaireet Métabolisme [PHYME], University of Geneva, Switzerland) for performing thehyperinsulinemic-euglycemic clamp, Jorge Altirriba Gutierrez (PHYME) for theLabMaster analysis, Clarissa Bartley (PHYME) for the b-cell extraction andcoculture, and Didier Collin (Geneva University Hospitals, Switzerland) for the PET/CTanalysis.Funding. This work was supported by a grant from Novartis, formerly Ciba-Geigy-Jubilee-Foundation; the Bo & Kerstin Hjelt Foundation; and the Sir Jules ThornCharitable Overseas Trust Reg. (to N.B).Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. N.B. conceived the study. M.T., P.M., M.F., and N.B.developed the methodology. J.B., F.B., and N.B. performed formal analysis andinvestigation. N.B. performed writing and original draft preparation. M.T., P.M., F.B.,M.F., and N.B. reviewed and edited the manuscript. N.B. acquired funding. N.B.,M.T., P.M., M.F., and N.B. provided resources. N.B. supervised the study. N.B. is theguarantor of this work and, as such, had full access to all the data in the studyand takes responsibility for the integrity of the data and the accuracy of the dataanalysis.

References1. Wan Y, Chong LW, Evans RM. PPAR-gamma regulates osteoclastogenesis inmice. Nat Med 2007;13:1496–15032. Kawai M, Sousa KM, MacDougald OA, Rosen CJ. The many facets of PPAR-gamma: novel insights for the skeleton. Am J Physiol Endocrinol Metab 2010;299:E3–E93. Cao J, Ou G, Yang N, et al. Impact of targeted PPARg disruption on boneremodeling. Mol Cell Endocrinol 2015;410:27–344. Sun H, Kim JK, Mortensen R, Mutyaba LP, Hankenson KD, Krebsbach PH.Osteoblast-targeted suppression of PPARg increases osteogenesis through activationof mTOR signaling. Stem Cells 2013;31:2183–21925. Fulzele K, Riddle RC, DiGirolamo DJ, et al. Insulin receptor signaling in oste-oblasts regulates postnatal bone acquisition and body composition. Cell 2010;142:309–3196. Ferron M, McKee MD, Levine RL, Ducy P, Karsenty G. Intermittent injections ofosteocalcin improve glucose metabolism and prevent type 2 diabetes in mice. Bone2012;50:568–575

diabetes.diabetesjournals.org Brun and Associates 2553

7. Wei J, Ferron M, Clarke CJ, et al. Bone-specific insulin resistance disruptswhole-body glucose homeostasis via decreased osteocalcin activation. J Clin Invest2014;124:1–138. Brennan-Speranza TC, Henneicke H, Gasparini SJ, et al. Osteoblasts mediatethe adverse effects of glucocorticoids on fuel metabolism. J Clin Invest 2012;122:4172–41899. Levinger I, Lin X, Zhang X, et al. The effects of muscle contraction and re-combinant osteocalcin on insulin sensitivity ex vivo. Osteoporos Int 2016;27:653–66310. Yoshikawa Y, Kode A, Xu L, et al. Genetic evidence points to an osteocalcin-independent influence of osteoblasts on energy metabolism. J Bone Miner Res 2011;26:2012–202511. Rodríguez-Carballo E, Gámez B, Méndez-Lucas A, et al. p38a function inosteoblasts influences adipose tissue homeostasis. FASEB J 2015;29:1414–142512. Mu Q, Fang X, Li X, et al. Ginsenoside Rb1 promotes browning through reg-ulation of PPARg in 3T3-L1 adipocytes. Biochem Biophys Res Commun 2015;466:530–53513. Jeremic N, Chaturvedi P, Tyagi SC. Browning of white fat: novel insight intofactors, mechanisms, and therapeutics. J Cell Physiol 2017;232:61–6814. Peyrou M, Bourgoin L, Poher AL, et al. Hepatic PTEN deficiency improvesmuscle insulin sensitivity and decreases adiposity in mice. J Hepatol 2015;62:421–42915. Braga M, Reddy ST, Vergnes L, et al. Follistatin promotes adipocyte differen-tiation, browning, and energy metabolism. J Lipid Res 2014;55:375–38416. Gamas L, Matafome P, Seiça R. Irisin and myonectin regulation in the insulinresistant muscle: implications to adipose tissue: muscle crosstalk. J Diabetes Res2015;2015:35915917. Wei J, Shimazu J, Makinistoglu MP, et al. Glucose uptake and Runx2 synergizeto orchestrate osteoblast differentiation and bone formation. Cell 2015;161:1576–159118. Bozec A, Bakiri L, Jimenez M, et al. Osteoblast-specific expression of Fra-2/AP-1controls adiponectin and osteocalcin expression and affects metabolism. J Cell Sci2013;126:5432–544019. Izawa T, Rohatgi N, Fukunaga T, et al. ASXL2 regulates glucose, lipid, anddkeletal homeostasis. Cell Reports 2015;11:1625–163720. Aarden EM, Burger EH, Nijweide PJ. Function of osteocytes in bone. J CellBiochem 1994;55:287–29921. Javaheri B, Stern AR, Lara N, et al. Deletion of a single b-catenin allele inosteocytes abolishes the bone anabolic response to loading. J Bone Miner Res 2014;29:705–71522. Thouverey C, Caverzasio J. Suppression of p38a MAPK signaling in osteoblastlineage cells impairs bone anabolic action of parathyroid hormone. J Bone Miner Res2016;31:985–99323. Trajkovski M, Hausser J, Soutschek J, et al. MicroRNAs 103 and 107 regulateinsulin sensitivity. Nature 2011;474:649–65324. Mouche S, Mkaddem SB, Wang W, et al. Reduced expression of the NADPHoxidase NOX4 is a hallmark of adipocyte differentiation. Biochim Biophys Acta 2007;1773:1015–102725. Mai G, Nguyen TH, Morel P, et al. Treatment of fulminant liver failure bytransplantation of microencapsulated primary or immortalized xenogeneic hepato-cytes. Xenotransplantation 2005;12:457–46426. Carobbio S, Ishihara H, Fernandez-Pascual S, Bartley C, Martin-Del-Rio R,Maechler P. Insulin secretion profiles are modified by overexpression of glutamatedehydrogenase in pancreatic islets. Diabetologia 2004;47:266–276

27. Bonnet N, Conway SJ, Ferrari SL. Regulation of beta catenin signaling andparathyroid hormone anabolic effects in bone by the matricellular protein periostin.Proc Natl Acad Sci U S A 2012;109:15048–1505328. Suárez-Zamorano N, Fabbiano S, Chevalier C, et al. Microbiota depletionpromotes browning of white adipose tissue and reduces obesity. Nat Med 2015;21:1497–150129. Marciano DP, Kuruvilla DS, Boregowda SV, et al. Pharmacological repression ofPPARg promotes osteogenesis. Nat Commun 2015;6:744330. Mieczkowska A, Baslé MF, Chappard D, Mabilleau G. Thiazolidinediones induceosteocyte apoptosis by a G protein-coupled receptor 40-dependent mechanism.J Biol Chem 2012;287:23517–2352631. Winkler DG, Sutherland MK, Geoghegan JC, et al. Osteocyte control of boneformation via sclerostin, a novel BMP antagonist. EMBO J 2003;22:6267–627632. Fu H, Desvergne B, Ferrari S, Bonnet N. Impaired musculoskeletal response toage and exercise in PPARb(-/-) diabetic mice. Endocrinology 2014;155:4686–469633. Lee NK, Sowa H, Hinoi E, et al. Endocrine regulation of energy metabolism bythe skeleton. Cell 2007;130:456–46934. Ferron M, Hinoi E, Karsenty G, Ducy P. Osteocalcin differentially regulates betacell and adipocyte gene expression and affects the development of metabolic dis-eases in wild-type mice. Proc Natl Acad Sci U S A 2008;105:5266–527035. Wei J, Hanna T, Suda N, Karsenty G, Ducy P. Osteocalcin promotes b-cellproliferation during development and adulthood through Gprc6a. Diabetes 2014;63:1021–103136. Frey J, Kim S, Li Z, et al. Sclerostin influences body composition by regulat-ing catabolic and anabolic metabolism in adipocytes. J Bone Miner Res 2016;31(Suppl. 1):103437. Tseng YH, Kokkotou E, Schulz TJ, et al. New role of bone morphogenetic protein7 in brown adipogenesis and energy expenditure. Nature 2008;454:1000–100438. Sugimoto H, Yang C, LeBleu VS, et al. BMP-7 functions as a novel hormone tofacilitate liver regeneration. FASEB J 2007;21:256–26439. Kinoshita K, Iimuro Y, Otogawa K, et al. Adenovirus-mediated expression ofBMP-7 suppresses the development of liver fibrosis in rats. Gut 2007;56:706–71440. Boon MR, van den Berg SA, Wang Y, et al. BMP7 activates brown adiposetissue and reduces diet-induced obesity only at subthermoneutrality. PLoS One 2013;8:e7408341. Wang L, Ciani C, Doty SB, Fritton SP. Delineating bone’s interstitial fluidpathway in vivo. Bone 2004;34:499–50942. Wang L, Wang Y, Han Y, et al. In situ measurement of solute transport in thebone lacunar-canalicular system. Proc Natl Acad Sci U S A 2005;102:11911–1191643. Price C, Zhou X, Li W, Wang L. Real-time measurement of solute transportwithin the lacunar-canalicular system of mechanically loaded bone: direct evidencefor load-induced fluid flow. J Bone Miner Res 2011;26:277–28544. Juffer P, Jaspers RT, Lips P, Bakker AD, Klein-Nulend J. Expression of muscleanabolic and metabolic factors in mechanically loaded MLO-Y4 osteocytes. AmJ Physiol Endocrinol Metab 2012;302:E389–E39545. Bellido T, Saini V, Pajevic PD. Effects of PTH on osteocyte function. Bone 2013;54:250–25746. Gorski JP, Huffman NT, Vallejo J, et al. Deletion of Mbtps1 (Pcsk8, S1p, Ski-1)gene in osteocytes stimulates soleus muscle regeneration and increased size andcontractile force with age. J Biol Chem 2016;291:4308–432247. Hevener AL, He W, Barak Y, et al. Muscle-specific Pparg deletion causes insulinresistance. Nat Med 2003;9:1491–149748. Lim J, Burclaff J, He G, Mills JC, Long F. Unintended targeting of Dmp1-Crereveals a critical role for Bmpr1a signaling in the gastrointestinal mesenchyme ofadult mice. Bone Res 2017;5:16049

2554 Bone Regulates Browning and Energy Metabolism Diabetes Volume 66, October 2017