Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/yexcr

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 1

0014-4827/$ - see frohttp://dx.doi.org/10.1

nCorresponding auFrance. Fax: þ33 957

E-mail address: y

Research Article

GW9508, a free fatty acid receptor agonist, specificallyinduces cell death in bone resorbing precursor cellsthrough increased oxidative stress frommitochondrial origin

Claire Philippea,b,c, Fabien Wauquiera,b,c, Laurent Léotoinga,b,c,Véronique Coxama,b,c, Yohann Wittranta,b,c,n

aINRA, UMR 1019, UNH, CRNH Auvergne, F-63009 Clermont-Ferrand, FrancebClermont Université, Université d′Auvergne, Unité de Nutrition Humaine, BP 10448, F-63000 Clermont-Ferrand, FrancecEquipe Alimentation, Squelette et Métabolismes, France

a r t i c l e i n f o r m a t i o n

Article Chronology:

Received 8 April 2013Received in revised form6 August 2013Accepted 7 August 2013Available online 22 August 2013

Keywords:

Free fatty acid receptor agonistBoneCell deathOxidative stress

nt matter & 2013 Elsevier016/j.yexcr.2013.08.013

thor at: Clermont Univers59 912.ohann.wittrant@clermont

a b s t r a c t

GW9508 is a free fatty acid receptor agonist able to protect from ovariectomy-induced bone loss in vivo

thought inhibition of osteoclast differentiation in a G-coupled Protein Receptor 40 (GPR40)-dependentway. In this study, we questioned whether higher doses of GW9508 may also influence resorbing cellviability specifically. Interestingly, GW9508 at 100 mM altered osteoclast precursor (OcP) viability whileit had positive effects on osteoblastic precursors suggesting an activity dependent on the cell lineage.According to 7-AAD/Annexin-V staining, induced OcP cell death was found to be associated withnecrosis mechanisms. Consistently, GW9508 led to a sustained establishment of oxidative stress frommitochondrial origin. In contrast to previous observations on osteoclast differentiation inhibition, OcPviability targeted by high doses of GW9508 appeared to be independent of GPR40 involvement.

Although mediating structures remain to be determined, our data demonstrate for the first time thatthis fatty acid receptor agonist driving OcP specific cell death may now open new perspectivesregarding therapeutic strategies in osteolytic disorders.

& 2013 Elsevier Inc. All rights reserved.

Introduction

Recently, orphan membrane G- protein coupled receptors (GPRCs)such as GPR40, 41, 43 and 120 [1] have been identified as fattyacid receptors. GPR41 and 43 are short chain fatty acid receptorswhile GPR40 and GPR120 preferentially bind medium/long chainfatty acids. GPR40 has mainly been studied in β-pancreatic cellfunctions for its role in potentiating glucose dependent insulin

Inc. All rights reserved.

ité, Université d'Auvergne,

.inra.fr (Y. Wittrant).

secretion in the presence of free fatty acids [2–5]. Latterly, GPR40functions were extended to brain tissues for its role in adultneurogenesis and memory mechanisms [6]. More generally,GPR40 play a pivotal role in nutrisensing along the gastro-intestinal tract, most notably for mediating fat related taste intong [7] and fat sensing in the intestine [8]. On the other hand,GPR120 was reported to mediate the anti-inflammatory effects ofω-3 polyunsaturated fatty acids (PUFA) in macrophages [9].

Unité de Nutrition Humaine, BP 10448, F-63000 Clermont-Ferrand,

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 13036

Regarding mineralized tissues, we and others previouslydemonstrated that GPR40 is also expressed in bone cells [10,11]and that GPR40�/� mice exhibit osteoporotic features [12]. Thisphenotype relies on a GPR40 protective effect towards bonematrix. Indeed, in vivo, GPR40 stimulation protects from boneloss in an ovariectomized mouse model through inhibition ofosteoclast differentiation. In vitro, we showed that doses as low as50 mM of the GPR40 agonist GW9508 targeted NF-κB pathway byblocking its transcriptional activity and subsequent osteoclastdifferentiation without affecting cell viability [13].However, some of these fatty acid receptors were recently

shown to differently impact cell behavior from several origins.GPR43 was proven to induce caspase activation and increaseapoptotic cell death upon propionate/butyrate treatment in HCT8human colonic adenocarcinoma cells [14]. Along with GPR43,Wu et al. demonstrated that GPR40 promote palmitate-inducedER stress and apoptosis in MIN6 β cells [15]. In contrast, GPR120activation was found to inhibit serum deprivation-induced apop-tosis in a murine enteroendocrine cell line STC-1 [16].In the light of these seemingly conflicting data, and regarding the

impact of pathological bone resorption in different health conditions,we questioned whether higher doses of GW9508 may influenceresorbing cell viability specifically and investigated occurring mechan-isms. In this study we demonstrate that high doses of GW9508 leadto an increase in pre-osteoclast (OcP) cell death through increasedoxidative stress in a cell origin dependent manner.Thus, according to our results, this fatty acid receptor agonist

driving OcP specific cell death could lead to new opportunities intherapeutic strategies specifically targeting osteolytic disordersincluding periprosthetic and tumors associated bone loss.

Materials and methods

Cell lines

The murine osteoclast and osteoblast precursors cell lines RAW264.7 and MC3T3-E1 were obtained from the American TypeCulture Collection (ATCC Numbers: TIB-71 and CRL-2593 respec-tively). For all experiments, cells were seeded at a density of3�104 cells/cm² and grown in α-MEM medium without phenolred supplemented with 10% FCS. At 80% confluency, cells wereserum starved for a total period of 48 hours for each conditionincluding GPR40 agonist incubation periods.

Cytotoxicity assay (XTT)

Cell cytotoxicity was determined by the XTT based method, usingthe Cell Proliferation Kit II (Roche) according to the supplier′srecommendations. As described above, cells were first serumstarved and then incubated for 24 h in the presence of vehicule(DMSO) or 1–100 mM of GW9508 (Cayman Chemical–CAS885101-89-3). Optical density was measured on a microplatereader using an absorbance wavelength of 450 nm, with areference wavelength of 650 nm.

Cell death: Annexin V/7-Amino-Actinomycin (7-AAD)staining

RAW 264.7 were cultured in the presence of vehicule (DMSO) or100 mM of GW9508 for increasing period of time ranging from 3

to 24 h prior cell staining. Staining was analyzed by fluorescentactivated cell sorting using the PE Annexin V Apoptosis DetectionKit I (BD Pharmingen) according to the supplier′s recommenda-tions. Annexin V staining was detected by phycoerythrin (PE)emission signal detector (FL2) and 7-AAD staining by FL3. 10,000positive events were used to estimate the distribution within thetotal population. Figures show representative data from threeindependent experiments.

Assay of caspase-3 activity

The CaspACE™ assay system (Promega) was used to quantifycaspase-3 activity in cell lysates following the manufacturer′sprotocol. Briefly, after 24 h of incubation with GW9508,RAW264.7 were washed with ice-cold PBS and lysed in ice-coldlysis buffer (NaCl 150 mM, Tris 50 mM, Nonidet P-40 1%, sodiumdeoxycholate 0.25%, NaF 1 mM, NaVO4 1 mM, leupeptine 10 μg/ml, aprotinin 10 μg/ml, PMSF 0.5 mM). For each treatment group,an equal amount of soluble protein was incubated with 50 μMacetyl-Asp-Glu-Val-Asp 7-amino-4-methyl coumarin (Promega),a fluorogenic substrate for caspase-3, with or without 50 μMacetyl-Asp-Glu-Val-Asp aldehyde, a specific caspase-3 inhibitor(Promega). Cell lysates were pre-incubated with the inhibitor for30 min prior to adding the substrate and cleavage of the substratewas analyzed using a fluorometer excitation wavelength 360 nm,detection wavelength 460 nm.

Measurement of reactive oxygen species (ROS) production(DCF-DA)

Intracellular reactive oxygen species (ROS) were detected using2′,7′-dichlorodihydrofluorescein diacetate (DCFDA) (Invitrogen).RAW 264.7 were cultured for 1, 2, 3 or 24 h either in the presenceof vehicule (DMSO) or 100 mM of GW9508. Cells were washedtwice with warm PBS and incubated for 30 min at 37 1C (in thedark) with the probe at a final concentration of 2.5 mM dye. Cellswere washed twice with PBS and harvested. The fluorescenceintensity was quantified by flow cytometry (FL1). 10,000 positiveevents were used to estimate the distribution within the totalpopulation. Figures show representative data from three inde-pendent experiments.

Measurement of mitochondrial ROS production (Mitosox)

Intracellular mitochondrial reactive oxygen species (ROS) produc-tion were detected using MitoSOX Red mitochondrial superoxideindicator for live-cell imaging (Invitrogen) according to thesupplier′s recommendations. RAW 264.7 were cultured for 24 hin the presence of vehicule (DMSO) or 100 mM of GW9508.The fluorescence intensity was quantified by flow cytometry(FL1). 10,000 positive events were used to estimate the distribu-tion within the total population. Figures show representative datafrom three independent experiments.

Measurement of Mitochondrial Membrane Potential (JC-1)

Mitochondrial membrane potential (MMP) was estimated using5,5′,6,6′-tetrachloro-1,1′,3,3′ tetraethylbenzimidazolylcarbocya-nine iodide (JC-1) (Invitrogen). As described previously, RAW264.7 were cultured for 24 h in the presence of vehicule (DMSO)

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 1 3037

or 100 mM of GW9508. Cells were washed twice with warm PBSand incubated for 10 min in the dark at 37 1C with the probe toprovide a final concentration of 10 mg/ml dye. Cells were washedtwice with PBS and harvested for analysis using flow cytometrymethods (FL1). 10,000 positive events were used to estimate thedistribution within the total population. Figures show represen-tative data from three independent experiments.

GPR40 silencingLong-term GPR40 knock-down in RAW264.7 cells was obtained usingthe SureSilencing™ shRNA Interference technology (SABiosciences,

RA

W26

4.7

cell

prol

ifera

tion

(fold

cha

nge)

0

0.2

0.4

0.6

0.8

1

1.2

1.4 *

DMSO 1 10 50 100

GW9508 (μM)

MC

3T3-

E1 c

ell p

rolif

erat

ion

(fold

cha

nge)

Annexin V

7-AADCel

l cou

nts

Fluorescence intensity

(2) GW9508 (100μM)(1) DMSO

(1) (2)

(1)(2)

(2) GW9508 (100μM)(1) DMSO

Ann

exin

V /

7 -A

AD

sta

inin

gPo

urce

ntag

e of

the

popu

latio

n

0

50

100

150

200

250

DMSO GW9508

Cas

pase

-3 a

ctiv

ity(%

ove

r DM

SO c

ontr

ol v

ehic

ule) *

Fig. 1 – GW9508 induces osteoclast precursor cell death. (A) XTT anacompared to control vehicule DMSO. (B) XTT analyzes in OcP cellsDMSO. (C) Fluorescence activating cell sorting data for 7-AAD andGW9508 (100 lM) or DMSO (control vehicule) for 24 h. (D) Correspothe total population [(þ) positive staining; (�) negative staining].(100 lM) expressed in pourcentage over the DMSO control vehicul

Qiagen). After one month selection, growing cells were harvested forGPR40 expression by real-time RT-PCR. Colonies exhibiting the high-est knock-down efficiency were selected for further investigations.

Western blot analysisCultured cells were immediately chilled on ice and lysed for30 min using NP40 lysis buffer. Proteins were quantified usingBCA kit (Sigma-Aldrich) according to the manufacturer′s protocoland 20 mg of total protein were loaded and resolved on a 10% SDS-PAGE gel and transferred to a nitrocellulose membrane (Invitro-gen). GPR40, and β-actin were detected using specific antibodies

0

0.2

0.4

0.6

0.8

1

1.2

*

GW9508 (μM)

DMSO 1 10 50 100

DMSO GW9508 0

20

40

60

80

100

120

Annexin V (+)7-AAD (+)

Annexin V (+)7-AAD (-)

Annexin V (-)7-AAD (+)

Annexin V (-)7-AAD (-)

lyzes in osteoblast precursor cells (MC3T3-E1) upon GW9508 as(RAW264.7) upon GW9508 as compared to control vehiculeAnnexin-V labeling in RAW264.7 cultures in the presence ofnding distribution of labeled cells expressed in pourcentage of(E) Caspase-3 activity in RAW264.7 upon GW9508 treatmente. (*); po0.05.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 13038

from Santa Cruz Biotechnology (Santa Cruz, CA; GPR40: sc-28416/β-actin: sc-47778). Proteins were revealed by chemiluminescenceusing ECL reagent (Amersham).

Statistical analysis

All experiments were performed in triplicate. Data obtained wereanalyzed by the non-parametric method Kruskal–Wallis (from twogroups or more) one-way analysis of variance. (*); (po0.05) indicatessignificant difference (ExcelStat Pro software–Microsoft Office 2007).

Results

High doses of GW9508 differentially impact proliferationof bone cells

To test whether high concentration of the GPR40 agonist maydifferentially impact bone cells proliferation, osteoclast and osteoblastprecursors, RAW264.7 and MC3T3-E1 cells respectively, were incu-bated with increasing doses of GW9508 ranging from 1 to 100 mM.XTT analyzes revealed that until 50 mM, GW9508 had no significanteffect on cell growth (Fig. 1A and B). However, while 100 mM slightly

0

20

40

60

80

100

unstained cellsstained cells

1h

Positive cells

Negative cells

(1)

(2)(3)

(4)

DC

FDA

sta

inin

g (R

OS

prod

uctio

n at

24

hour

s)

DMSO GW9508 (100μM)

Cel

l cou

nts

DCF-DA fluorescenc

(1): No probe; (2): DMSO; (3): GW9508 (100μM); (4): positi

Fig. 2 – Oxidative stress establishment. (A) Reactive oxygen specieRAW264.7 cells were incubated for 1 h to 3 h with either GW9508 (as positive control). (B) Reactive oxygen production as expressed iincubation in the presence of GW9508 (100 lM) or DMSO.

increased pre-osteoblast proliferation, a drop of 50% absorbancevalues in OcP cultures was observed demonstrating that higherconcentration of GW9508 specifically counters growth of pro-resorbing cells (Fig. 1B; 58.7%78.7 versus 100%75.4 in control-treated cells).

Inhibition of osteoclast precursor proliferation isassociated with enhanced cell death markers

To determine whether cell growth inhibition was related tocytostatic effects or cell death, OcP were labeled for apoptosisand necrosis pathways as determined by annexin V/7-AADapoptosis detection kit. According to fluorescent activating cellsorting data, when incubated with 100 mM of GW9508, cells werefound positive for both markers (63.1% at 24 h treatment versus22.1% in control-treated cells) (Fig. 1C and D). This double annexinV/7-AAD labeling occurred even at early time points (Fig. 5;supplemental data) suggesting that high doses of GW9508 mayinduce preferentially necrosis pathways. However, caspase-3activity, a pivotal mediator in apoptosis processes was signifi-cantly higher after treatment with 100 mM of GW9508 for 24 hwhen compared to control-treated cells (Fig. 1E), thus uncoveringa probable more complex process.

2h 3h

e intensity (FL1-H)

ve control (Tert-butyl 20μM)

s production as measured by FACS using the DCF-DA probe.100 lM), DMSO (control vehicule) or hydroxyl tert-butyl (20 lMn pourcentage of RAWS264.7 DFC-DA positive cells after 24 h

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 1 3039

GW9508-induced necrosis is related to oxidative stressestablishment in osteoclast precursors

In order to decipher downstream mechanisms, oxidative stressestablishment was explored. Using DCF-DA, cells were evidencedfor a GW9508 dependent stimulation of reactive oxygen speciesproduction. Investigating time course of stress establishment,data from FACS analyzes revealed that ROS production startedfrom 1 hour as compared to vehicule control DMSO and lasteduntil 24 h incubation (Fig. 2A and B) thus supporting necrosismechanisms. A dose-response was also determined (Fig. 6;supplemental data).

Oxidative stress originates from the mitochondria

During cell death, oxidative stress mainly results from an electronleak in the respiratory chain. To further elucidate the origin of theGW9508-induced ROS production, OcP were tested for specificmitochondrial ROS production using mitosox technologies(Fig. 3A and B). Data reveal that GW9508 induces a markedrelease of superoxide anions from the mitochondria after 1 hincubation with GW9508. This release not only lasted but alsoincreased until 24 h of treatment (Fig. 3B). In addition, consis-tently with the mitosox data, JC-1 analyzes demonstrate a partialloss of the mitochondrial potential (Fig. 3C) thus strongly

DMS

JC-1

red

fluor

esce

nce

(1)(2)

Cel

l cou

nts

MitoSOX Fluorescence intensity (FL2-H)

(1): DMSO(2): GW9508 (100μM)

24h

1h

(1)

(2)

Cel

l cou

nts

MitoSOX -Fluorescence in

(1): DMSO(2): GW9508 (10

(1): DMSO;(2): GW9508 (100μM)

(1)

Fig.3 – Mitochondrial contribution to oxidative stress establishmemitochondria analyzed by fluorescence activating cell sorting usingwith either GW9508 (100 lM) or DMSO (control vehicule). (C) Mitoprobe. As described previously, RAW 264.7 were cultured for 24 hCarbonyl cyanide 3-chlorophenylhydrazone (CCCP) was used as po

supporting the involvement of the mitochondria in the oxidativestress establishment and the subsequent cell necrosis.

Alteration of osteoclast precursor viability uponincubation with high doses of GW9508 is not dependent onthe presence of GPR40.

To conclude on the mechanisms of action, we investigated to roleof the most specific GW9508 binding receptor: GPR40. Todecipher its role in mediating GW9508-induced OcP cell death,GPR40 was knockdowned in RAW264.7 cells using shRNA tech-nologies. Scramble shRNA was used as control (Fig. 4A). As shownin Fig. 4B, incubation with 100 mM of GW9508 decreased by halfcell viability in unstransfected or scrambled cells as expected(Fig. 1). Surprisingly, the absence of GPR40 had no impact on theGW9508 effects.

Discussion

Our results show for the first time that high doses of a free fattyacid receptor agonist: GW9508 plays a crucial role in boneresorbing cell behavior in a cell-origin dependent manner. Thisfree fatty acid receptor agonist-induced cell death was revealedby an early joined increase in both, annexinV and 7-AAD staining,

CCCPO GW9508 (100μM)

JC-1 green fluorescence

2h 3h

tensity (FL2-H)

(1): DMSO(2): GW9508 (100μM)0μM)

(2)(1)

(2)

nt. (A) and (B) Specific superoxide anion production byMitoSOX probe. RAW264.7 cells were incubated for 1 h to 24 hchondrial membrane potential (MMP) estimated using JC-1in the presence of vehicule (DMSO) or 100 lM of GW9508.sitive control.

GPR 40

β-Actine

00.20.40.60.8

11.21.4

scr40

GPR40

*

GW9508 (μM)

*

RA

W26

4.7

cell

prol

ifera

tion

(OD

.10-3

.min

-1)

Rel

ativ

e m

RN

A

expr

essi

on

sh40

Fig. 4 – Role of GPR40 in GW9508-induced OcP cell death. (A) RT-PCR and Western blot analyzes of GPR40 knockdown efficiency inRAW264.7 cells. (B) XTT analyzes in OcP cells (unstransfected RAW264.7: right bars; transfected RAW264.7 cells with scrambleshRNA: middle bars; transfected RAW264.7 cells with GPR40 shRNA: left bars) upon GW9508 incubation as compared to controlvehicule DMSO. (*); po0.05 GW9508 100 lM versus control DMSO.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 13040

associated with a strong oxidative stress establishment, thussupporting a necrosis related process.In this study, GW9508 induced-cell death parallels with the

lipoapoptosis observed in beta cells upon prolonged GPR40activation by palmitate [17]. In beta cells, lipoapoptosis requiredoxidative stress establishment as demonstrated by NO production[5,18]. In our hands, GW9508-dependent OcP cell death wasassociated with early reactive oxygen species production. Thisoxidative stress may result from a wide variety of sources [19].Therefore, cells were tested for NO production, NADPHoxidasesand xanthine oxidase activities. Experiments' data revealed anabsence of significant variations in the presence of GW9805 foreither one of these sources (Fig. 7; supplemental data). Incontrast, our results evidence a great participation of mitochon-dria as demonstrated by the increase in specific superoxide anionproduction and loss of mitochondrial potential. These resultswere strengthened by the measurements of mitochondrial com-plexes activities revealing an increase in ROS production from therespiratory chain (data not shown). This ROS production frommitochondrial origin strongly supports the induction of necroticmechanisms [20]. Nevertheless, the significant raise in caspase-3activity upon GW9508 suggests that both necrotic and apoptosisprocesses may be involved [21]. More precisely, this OcP celldeath induced by a well identified GPR40 agonist refers to areceptor dependent necrosis, named necroptosis [22]. To furtherinvestigate this pathway, RIP complex activation, a pivotalnecroptosis molecular crossroad [23,24], was explored. Still, datarevealed that incubation with 100 mM of GW9508 increased RIPexpression without significant activation, thus requiring deeperinvestigation (data not shown).Regarding mediating structures, experiments done in GPR40

knockdown cells reveal that osteoclast precursors were notprotected from GW9508 impact on cell viability in the absenceof GPR40. GW9508 presents a higher affinity for the receptorGPR40. However, GW9508 also activates GPR120 in a lesser extent[25]. Thus GPR120 could be responsible for the effect on cellviability that only occurred at high doses. Together with ourprevious results showing that GW9508 inhibits OcP cell differ-entiation through a GPR40 dependent mechanism for low doses

of GW9508, one may speculate that high doses of GW9508 may infact induce bone resorbing cell death through GRP120 activation.

The most disturbing data from this study lays in the cell originspecificity: why does GW9508 target bone resorbing cells while incontrast osteoblastic cell proliferation tends to be stimulated,even though both cell lines are from murine origin? Part of theexplanation could reside in the expression level of each receptor.However, we and others [11] found that GPR40 expression levelwas higher in osteoclast than in osteoblast while the opposite wasobserved for GPR120. Since, GW9508 tends to stimulate osteo-blast proliferation, further investigations are needed to bringanswers.

Finally, although mediating receptor remains to be deciphered,our data open new avenues in terms of clinical investigationsregarding the potential benefit of fatty acid receptor agonists onbone tissue and lead to new opportunities in the design oftherapeutic strategies targeting pathological osteolysis includingprosthetic bone loss and most notably tumor associated osteolysis.

Acknowledgments/Disclosure

None of the authors have a financial interest related to thismanuscript.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.yexcr.2013.08.013.

r e f e r e n c e s

[1] C.P. Briscoe, et al., The orphan G protein-coupled receptor GPR40is activated by medium and long chain fatty acids, J. Biol. Chem.278 (13) (2003) 11303–11311.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 9 ( 2 0 1 3 ) 3 0 3 5 – 3 0 4 1 3041

[2] M. Kebede, et al., The fatty acid receptor GPR40 plays a role ininsulin secretion in vivo after high-fat feeding, Diabetes 57 (9)(2008) 2432–2437.

[3] P. Newsholme, et al., Life and death decisions of the pancreatic beta-cell: the role of fatty acids, Clin Sci. London 112 (1) (2007) 27–42.

[4] P. Steneberg, et al., The FFA receptor GPR40 links hyperinsuli-nemia, hepatic steatosis, and impaired glucose homeostasis inmouse, Cell Metab. 1 (4) (2005) 245–258.

[5] Y. Zhang, et al., The role of G protein-coupled receptor 40 inlipoapoptosis in mouse beta-cell line NIT-1, J. Mol. Endocrinol. 38(6) (2007) 651–661.

[6] T. Yamashima, ‘PUFA-GPR40-CREB signaling’ hypothesis for theadult primate neurogenesis, Prog. Lipid. Res. 51 (3) (2012) 221–231.

[7] C. Cartoni, et al., Taste preference for fatty acids is mediated byGPR40 and GPR120, J. Neurosci. 30 (25) (2010) 8376–8382.

[8] A.P. Liou, et al., The G-protein-coupled receptor GPR40 directlymediates long-chain fatty acid-induced secretion of cholecysto-kinin, Gastroenterology 140 (3) (2011) 903–912.

[9] D.Y. Oh, et al., GPR120 is an omega-3 fatty acid receptormediating potent anti-inflammatory and insulin-sensitizingeffects, Cell 142 (5) (2010) 687–698.

[10] J. Cornish, et al., Modulation of osteoclastogenesis by fatty acids,Endocrinology 149 (11) (2008) 5688–5695.

[11] A. Mieczkowska, et al., Thiazolidinediones induce osteocyteapoptosis by a G protein-coupled receptor 40-dependentmechanism, J. Biol. Chem. 287 (28) (2012) 23517–23526.

[12] F. Wauquier, C. Philippe, L. Léotoing, P. Lebecque, S. Mercier, M. J.Davicco, P. Pilet, J. Guicheux, E. Miot-Noirault, T. Alquier, V.Poitout, V. Coxam, Y. Wittrant, GPR40, a free fatty acid receptorinvolved in osteoporosis prevention, in: Proceedings of theAmerican Society for Bone and Mineral Research, San Diego,2011.

[13] F. Wauquier, et al., The free fatty acid receptor g protein-coupledreceptor 40 (GPR40) protects from bone loss through inhibition ofosteoclast differentiation, J. Biol. Chem. 288 (9) (2013) 6542–6551.

[14] Y. Tang, et al., G-protein-coupled receptor for short-chain fatty acidssuppresses colon cancer, Int. J. Cancer 128 (4) (2011) 847–856.

[15] J. Wu, et al., Inhibition of GPR40 protects MIN6 beta cells frompalmitate-induced ER stress and apoptosis, J. Cell Biochem. 113(4) (2012) 1152–1158.

[16] S. Katsuma, et al., Free fatty acids inhibit serum deprivation-induced apoptosis through GPR120 in a murine enteroendocrinecell line STC-1, J. Biol. Chem. 280 (20) (2005) 19507–19515.

[17] M. Elsner, W. Gehrmann, S. Lenzen, Peroxisome-generatedhydrogen peroxide as important mediator of lipotoxicity ininsulin-producing cells, Diabetes 60 (1) (2011) 200–208.

[18] S. Meidute Abaraviciene, et al., Palmitate-induced beta-celldysfunction is associated with excessive NO production and isreversed by thiazolidinedione-mediated inhibition of GPR40transduction mechanisms, PLoS One 3 (5) (2008) e2182.

[19] W. Droge, Free radicals in the physiological control of cellfunction, Physiol. Rev. 82 (1) (2002) 47–95.

[20] N. Apostolova, A. Blas-Garcia, J.V. Esplugues, Mitochondria sen-tencing about cellular life and death: a matter of oxidative stress,Curr. Pharm. Des. 17 (36) (2011) 4047–4060.

[21] S. Tanaka, et al., Regulation of osteoclast apoptosis by Bcl-2family protein Bim and Caspase-3, Adv. Exp. Med. Biol. 658(2010) 111–116.

[22] D.E. Christofferson, J. Yuan, Necroptosis as an alternative form ofprogrammed cell death, Curr. Opin. Cell Biol. 22 (2) (2010) 263–268.

[23] V. Temkin, M. Karin, From death receptor to reactive oxygenspecies and c-Jun N-terminal protein kinase: the receptor-interacting protein 1 odyssey, Immunol. Rev. 220 (2007) 8–21.

[24] P. Vandenabeele, et al., The role of the kinases RIP1 and RIP3 inTNF-induced necrosis, Sci. Signal 3 (115) (2010) re4.

[25] C.P. Briscoe, et al., Pharmacological regulation of insulin secretionin MIN6 cells through the fatty acid receptor GPR40: identifica-tion of agonist and antagonist small molecules, Br. J. Pharmacol.148 (5) (2006) 619–628.