FXR activation normalizes insulin sensitivity in visceral preadipocytes of a rabbit model of MetS

JournalofEndocrinology

ResearchE MANESCHI, L VIGNOZZI and others FXR signaling improves adipose

tissue function218 :2 215–231

FXR activation normalizes insulinsensitivity in visceral preadipocytesof a rabbit model of MetS

Elena Maneschi*, Linda Vignozzi*, Annamaria Morelli1, Tommaso Mello2,

Sandra Filippi3, Ilaria Cellai, Paolo Comeglio, Erica Sarchielli1, Alessandra Calcagno4,

Benedetta Mazzanti5, Roberto Vettor4, Gabriella Barbara Vannelli1, Luciano Adorini6

and Mario Maggi

Sexual Medicine and Andrology Unit, Department of Experimental and Clinical Biomedical Sciences,

University of Florence, Viale Pieraccini 6, Florence, Italy1Department of Experimental and Clinical Medicine, 2Gastroenterology Unit, Department of Experimental and

Clinical Biomedical Sciences and 3Interdepartmental Laboratory of Functional and Cellular Pharmacology of

Reproduction, Department of Neuroscience, Drug Research and Child Care, University of Florence, Florence, Italy4Department of Medicine, University of Padua, Padua, Italy5Hematology Unit, Department of Experimental and Clinical Medicine, University of Florence, Florence, Italy6Intercept Pharmaceuticals, 18 Desbrosses Street, New York, New York 10013, USA

*(E Maneschi and L Vignozzi contributed equally to this work)

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

Published by Bioscientifica Ltd.

Correspondence

should be addressed

to M Maggi

Email

[email protected]

Abstract

Insulin resistance is the putative key underlying mechanism linking adipose tissue (AT)

dysfunction with liver inflammation and steatosis in metabolic syndrome (MetS). We have

recently demonstrated that the selective farnesoid X receptor (FXR) agonist obeticholic acid

(OCA) ameliorates insulin resistance and the metabolic profile with a marked reduction in the

amountof visceralAT (VAT) in ahigh-fatdiet (HFD)-induced rabbitmodelofMetS. These effects

weremediatedby theactivationofFXR, since treatmentwith the selectiveTGR5agonist INT-777

wasnotable toameliorate themetabolicparameters evaluated.Herein,we report theeffectsof

in vivoOCA dosing on the liver, the VAT, and the adipogenic capacity of VAT preadipocytes

(rPADs) isolated from rabbits on a HFD compared with those on a control diet. VATand liver

were studiedby immunohistochemistry,Westernblotanalysis, andRT-PCR. rPADswereexposed

to a differentiating mixture to evaluate adipogenesis. Adipocyte size, hypoxia, and the

expressionofperilipinandcytosolic insulin-regulatedglucose transporterGLUT4 (SLC2A4)were

significantly increased in VAT isolated from the HFD rabbits, and normalized by OCA. The

expression of steatosis and inflammationmarkers was increased in the liver of the HFD rabbits

and normalized by OCA. rPADs isolated from the HFD rabbits were less sensitive to insulin, as

demonstrated by the decreased insulin-induced glucose uptake, triglyceride synthesis, and

adipogenic capacity, as well as by the impaired fusion of lipid droplets. OCA treatment

preserved all the aforementioned metabolic functions. In conclusion, OCA dosing in a MetS

rabbitmodel ameliorates liver andVAT functions. This could reflect the abilityofOCAto restore

insulin sensitivity in ATunable to finalize its storage function, counteracting MetS-induced

metabolic alterations and pathological AT deposition.

Key Words

" adipogenesis

" lipid droplet

" insulin signaling

" glucose transport

" preadipocytes

" liver

" steatohepatitis

Journal of Endocrinology

(2013) 218, 215–231

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 216

Introduction

The metabolic syndrome (MetS) is a constellation of

metabolic abnormalities centered on insulin resistance

and visceral fat accumulation. In addition to insulin

resistance and visceral obesity, MetS is accompanied by

dyslipidemia with high triglyceride levels and low HDL

cholesterol concentrations and hypertension (Despres &

Lemieux 2006). Non-alcoholic fatty liver disease (NAFLD),

a pathophysiological accumulation of lipids in the liver, is

considered the hepatic hallmark of MetS (Farrell 2009).

NAFLD, when associated with inflammation, evolves into

non-alcoholic steatohepatitis (NASH), which can lead to

cirrhosis and hepatocarcinoma. Insulin resistance has

been implicated in both the initiation of NAFLD and its

transition into NASH (Larter et al. 2010).

In MetS, adipose tissue (AT) is not only increased in

mass, but also dysfunctional and characterized by hyper-

trophic insulin-resistant adipocytes, fulfilling their storage

function but unable to take up any more triglycerides. The

excess of circulating triglycerides ultimately leads to fat

accumulation in ectopic areas, such as the liver, heart, and

skeletal muscle. This ectopic fat deposition amplifies insulin

resistance and can interfere with cellular functions (Despres

& Lemieux 2006, Virtue & Vidal-Puig 2010, Snel et al. 2012).

In response to triglyceride overload, naive preadipocytes

might be prompted to differentiate into mature adipocytes,

thus serving as a buffer against lipid accumulation in non-

adipose cells. However, insulin resistance may impair the

differentiation of preadipocytes, with a consequent inability

to store excess lipids by enlarged mature adipocytes

(Gustafson et al. 2009). Finally, the presence of insulin-

resistant adipocytes is considered as the key distinguishing

feature between ‘metabolically healthy’ and ‘metabolically

abnormal’ obesity (Samocha-Bonet et al. 2012).

During the past few years, bile acids (BAs) have

emerged as important modulators of metabolic homeo-

stasis and insulin resistance (Thomas et al. 2008). BAs

through dedicated receptors – in particular the nuclear

hormone receptor farnesoid X receptor (FXR, also known

as NR1H4) and the G protein-coupled receptor TGR5 –

modulate several metabolic pathways regulating glucose,

triglyceride, and cholesterol levels and energy homeo-

stasis. Interestingly, Fxr-deficient (FxrK/K) mice display

elevated plasma and hepatic cholesterol and triglyceride

levels, along with an accelerated hepatic response on

being fed a high-carbohydrate diet, and develop periph-

eral insulin resistance (Sinal et al. 2000). Three indepen-

dent reports have linked FXR deficiency to impaired

insulin sensitivity (Cariou et al. 2006, Ma et al. 2006,

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

Zhang et al. 2006). In addition, Tgr5 (Gpbar1)-deficient

mice exhibit impaired glucose tolerance (Thomas et al.

2009). In mice fed a high-fat diet (HFD), INT-777, a specific

TGR5 agonist without FXR agonist activity (Thomas et al.

2009) lowers serum glucose and insulin levels and

improves glucose tolerance (Sato et al. 2007).

The most clinically advanced FXR agonist is the

semi-synthetic BA derivative obeticholic acid (OCA,

6-ethyl-chenodeoxycholic acid or INT-747), which is able

to improve insulin sensitivity in patients with type 2

diabetes and NAFLD (Adorini et al. 2012, Mudaliar et al.

2013). OCA is a first-in-class FXR agonist, endowed with

high binding affinity and potency for FXR (EC50 0.1 mM),

with a 200-fold lower activity for TGR5 (Rizzo et al. 2010).

We have recently developed a non-genomic model of

MetS, by feeding rabbits a HFD. Such a model recapitulates

the human MetS phenotype (hypertension, hyper-

glycemia, dyslipidemia, VAT accumulation, and glucose

intolerance), including a condition of hypogonadotrophic

hypogonadism, as we have demonstrated in several

previous studies (Filippi et al. 2009, Vignozzi et al. 2011,

2012, Maneschi et al. 2012, Morelli et al. 2012, 2013).

Interestingly, VAT isolated from MetS animals was also

dysfunctional, being characterized by insulin-resistant

preadipocytes with impaired triglyceride synthesis and

adipogenesis (Maneschi et al. 2012). OCA dosing in MetS

rabbits not only prevents HFD-induced VAT expansion,

but also reduces fasting glucose levels and glucose

intolerance (Vignozzi et al. 2011, Morelli et al. 2012). In

addition, OCA treatment ameliorates MetS-associated

dysfunctions in corpora cavernosa (Vignozzi et al. 2011)

and bladder (Morelli et al. 2012).

The aim of the present study was to investigate the

role played by OCA in VAT dysfunction and steatohepa-

titis not only by evaluating the morphological and

functional features of the liver and VAT, but also by

analyzing the insulin sensitivity of VAT preadipocytes. In

particular, the insulin response of rabbit preadipocytes,

isolated from the different experimental groups, was

investigated in terms of triglyceride synthesis and lipid

droplet formation and mRNA expression of adipogenesis-

specific genes, as well as glucose uptake. We also report the

effect of treatment with the selective TGR5 agonist

INT-777 to discriminate between FXR- and TGR5-mediated

metabolic activities. The rabbit model of HFD-induced

MetS allowed us to analyze in detail AT function, providing

sufficient amount of visceral fat for all the different

experimental purposes.

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 217

Our results indicate that OCA treatment ameliorates,

via FXR activation, liver and VAT functions, most

probably by restoring insulin sensitivity in VAT.

Materials and methods

MetS rabbit model

The HFD-induced rabbit model of MetS was obtained as

described previously (Filippi et al. 2009). Male New

Zealand White rabbits (Charles River, Calco, Lecco,

Italy), weighing about 3 kg, were randomly numbered

and assigned to two different groups: untreated group

(nZ38), fed a control diet (CON), or treated group (nZ60),

fed a HFD (0.5% cholesterol and 4% peanut oil), for

12 weeks. The composition of the CON and HFD is reported

in Table 1. A subgroup of HFD rabbits was planned to be

treated with the FXR and TGR5 agonist OCA (10 mg/kg,

daily 5 days a week for 12 weeks, by oral gavage; nZ18), as

described previously (Vignozzi et al. 2011, Morelli et al.

2012), or with the selective TGR5 agonist INT-777

(30 mg/kg, daily 5 days a week for 12 weeks, by oral gavage;

nZ6; Pellicciari et al. 2009). The dose of OCA used was

selected based on the efficacy and pharmacokinetics

analysis carried out in rodents (Pellicciari et al. 2002). After

a 3-month chronic feeding at the dose of 10 mg/day per kg

BW to the rabbits, OCA was mainly present in the plasma as

a glycine conjugate (20% of the total BAs) and in lower

amounts (15%) as the unconjugated parent compound. No

other major metabolites resulting from a 7-dehydroxylation

process, glucuronides, and other polar metabolites were

identified (Intercept Pharmaceuticals (New York, NY, USA),

Internal Report 2011). Structure, potency, and selectivity

toward other nuclear hormone receptors have been

described previously (Pellicciari et al. 2002, Rizzo et al.

2010). Blood samples were obtained from marginal ear vein

at baseline and at week 12 in all the groups. Mean arterial

pressure measurements and oral glucose tolerance test

were carried out before killing, as described previously

Table 1 Composition of the control diet and high-fat diet

Analysis

Control diet

(CON) (%)

High-fat diet

(HFD) (%)

Water 12 12Protein 16.5 12.6FatVegetable derived 3.5 6Animal derived 0 0.5Fiber 15.5 21.2Ash 8.5 9.2

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

(Filippi et al. 2009). After 12 weeks of treatment, the rabbits

were killed using a lethal dose of pentobarbital (100 mg/kg),

and the specimens of the liver, VAT (accumulated between

the intestinal loops and mesentery), and gallbladder were

carefully excised, weighed, collected, and processed for the

subsequent analyses. VAT samples from all the rabbit groups

were also processed for the isolation of preadipocytes.

Biochemical and hormonal serum analyses were performed

as described previously (Filippi et al. 2009, Morelli et al.

2012, Vignozzi et al. 2012). Based on an interim analysis,

due to the lack of an effect of INT-777 on hyperglycemia and

on overall MetS parameters (see below) and to an

unexpected gallbladder hypertrophy, experiments with

INT-777 were stopped, and therefore available data are

limited to six rabbits.

To evaluate the effects of MetS, we designed an

algorithm taking into account the presence, as a dummy

variable, of one or more of the following factors: hyper-

glycemia, high triglyceride levels, high cholesterol levels,

increased blood pressure, and visceral fat accumulation.

Cut-offs for each factor were derived by the meanG2 S.D.

of the analyzed parameter, as measured in the CON

rabbits. Positivity for three or more factors indicates MetS.

Ethics statement

This study was carried out in strict accordance with the

recommendations in the Italian Ministerial Law #116/92

for the Care and Use of Laboratory Animals. The protocol

was approved by the Institutional Animal Care and Use

Committee of the University of Florence (protocol

number: 4/III). All surgery was performed under sodium

pentobarbital anesthesia, and all efforts were made to

minimize suffering.

Sample size

Assuming a probability of the occurrence of MetS of 2.5%

in the CON group and a probability of 60% in the HFD

group (data derived from our previous publications on the

same model (Filippi et al. 2009, Vignozzi et al. 2011, 2012,

Maneschi et al. 2012, Morelli et al. 2012, 2013)), the use of

74 rabbits with an allocation ratio of 1:1 between the

groups allows a power close to 100% in distinguishing a

difference in the rate of development of MetS between the

two treatment groups. Assuming a probability of the

occurrence of MetS equal to 60% in the group fed the HFD

and a probability of 10% in the group fed the HFDCOCA

(data derived from our previous publications on the same

model (Vignozzi et al. 2011, Morelli et al. 2012)), the use of

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 218

54 rabbits with an allocation ratio of 2:1 allows a power of

about 95% in distinguishing a difference in the rate of

development of MetS between the two treatment groups.

Histomorphometric analysis of VAT

VAT specimens were analyzed by hematoxylin and eosin

staining to measure adipocyte diameter, as described

previously (Maneschi et al. 2012), using the Nikon

Microphot-FXA microscope (Nikon, Tokyo, Japan) equip-

ped with the free software program ImageJ (NIH, Bethesda,

MD, USA), considering adipocytes to be regularly spherical.

Hypoxia detection and immunohistochemistry

VAT oxygenation was analyzed using the bio-reductive

drug pimonidazole hydrochloride (hypoxyprobe-1,

60 mg/kg), injected i.p. 1 h before killing, as described

previously (Maneschi et al. 2012, Morelli et al. 2012, 2013,

Vignozzi et al. 2012).

Preparation of total and membrane/cytosolic fractions for

western blot analysis

For protein extraction from the VAT samples, the frozen

tissues were ground in liquid nitrogen and divided into

two aliquots: one for total protein extraction and the other

for membrane/cytosolic preparations. Membrane and

cytosolic fractions were prepared using the ProteoExtract

subcellular proteome extraction kit (Calbiochem-Merck

KGaA, Darmstadt, Germany), according to the manufac-

turer’s instructions. Protein extracts were quantified with

the BCA reagent (Pierce, Rockford, IL, USA), and 15 mg of

each sample were resolved by 10% SDS–PAGE. Western

blot analysis with an anti-glucose transporter type 4 (GLUT4)

antibody (Upstate Biotechnology, Lake Placid, NY, USA) and

anti-perilipin antibody (Santa Cruz Biotechnology, Inc.) was

performed as described previously (Maneschi et al. 2012).

Equal protein loading was verified by reprobing the

membrane with an anti-actin antibody (Santa Cruz Bio-

technology, Inc.). Densitometry analysis of band intensity

was performed using the Photoshop 5.5 Software (Adobe

Systems, Inc. Italia srl).

Liver histology

Liver steatosis was assessed by Oil Red O staining of the

liver sections. Frozen sections were cut in a cryostat and

fixed in 4% paraformaldehyde for 20 min at room

temperature (RT). Then, the sections were treated for

2–5 min with isopropanol and stained with Oil Red O for

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

20 min. Oil Red O was prepared by diluting a stock

solution (0.3 g of Oil Red O in 100 ml of isopropanol) with

water (3:2) followed by filtration. After Oil Red O staining,

the sections were washed several times in water and

stained with hematoxylin and eosin to highlight the

hepatocyte nuclei. Finally, the sections were photo-

graphed, and computer-assisted quantification of Oil Red

O positivity was done after background subtraction using

the Adobe Photoshop 6.0 Software (Adobe Systems).

Immunohistochemistry for TNFa (TNF) in the liver sections

Liver sections were incubated overnight at 48C with a

primary anti-TNFa (TNF) antibody (infliximab 1:100

vol/vol, DakoCytomation, Copenhagen, Denmark). The

sections were rinsed in PBS and incubated with a

biotinylated secondary antibody and then with a strepta-

vidin–biotin–peroxidase complex (Ultravision large

volume detection system anti-polyvalent, Lab Vision,

Fremont, CA, USA). The reaction product was developed

with 3 0,3 0-diaminobenzidine tetrahydrochloride as the

chromogen (Sigma–Aldrich). Control experiments were

performed by omitting the primary antibody. The slides

were evaluated and photographed using a Nikon Micro-

phot-FXA microscope. Computer-assisted quantification

of the staining of TNFa was done after background

subtraction using the Adobe Photoshop 6.0 Software

(Adobe Systems).

Isolation, characterization, and differentiation of

rabbit visceral fat preadipocytes

The isolation of rabbit preadipocytes (rPADs) from VAT

was carried out as described previously (Maneschi et al.

2012). Briefly, VAT samples were digested with 1 mg/ml

collagenase type 2 (Sigma–Aldrich) for 1 h, treated with

red blood cell lysis buffer (155 mM NH4Cl, 10 mM KHCO3,

and 0.1 mM EDTA; 10 min at RT), then centrifuged at

2000 g for 10 min at RT, resuspended in a complete

medium (DMEM containing 10% fetal bovine serum

(FBS), 100 mg/ml streptomycin, 100 U/ml penicillin,

2 mM L-glutamine, and 1 mg/ml amphotericin-B; Sigma–

Aldrich), and filtered through a 150 mm mesh filter to

remove debris. Finally, the cells were cultured in a

complete culture medium at 378C in a humidified

atmosphere of 95% air–5% CO2. A subconfluent (90% of

the cell culture dish) and homogeneous fibroblast-like cell

population at passage 0 (P0) was obtained after 4–5 days

of culture. The subconfluent cells were trypsinized and

plated in cell culture dishes (P1). For all the experiments,

only P1 cultures were used, and the experiments were

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 219

repeated using at least three different rPAD preparations

for each experimental group. rPADs were characterized

by flow cytometry with the following conjugated mono-

clonal antibodies: CD34-PE, CD45-FITC, CD31-FITC,

CD14-PE, CD90-PE, CD106-FITC (BD Pharmingen, San

Diego, CA, USA), and CD105 PE (Ancell, Bayport, MN,

USA), as described previously (Maneschi et al. 2012). The

differentiation of rPADs, 2 days after confluence (time 0),

was induced by exposing them to a differentiation mixture

(DIM) containing 5 mg/ml insulin, 1 mM dexamethasone,

and 0.5 mM 3-isobutyl-1-methylxanthine (IBMX) in 5%

stripped FBS-supplemented DMEM for 8 days (Student

et al. 1980). The culture medium was replaced every 48 h,

and then the cells were shifted into a medium containing

5 mg/ml insulin for 48 h.

Qualitative and quantitative estimation of

triglyceride accumulation

Qualitative and quantitative analyses of intracellular lipids

were carried out using Oil Red O staining (Sigma) and

AdipoRed Assay (Cambrex BioScience, Walkersville, MD,

USA) respectively, as described previously (Maneschi et al.

2012). Briefly, both untreated and DIM-induced rPADs

were washed in PBS and fixed in 10% formalin for 1 h at

RT, followed by staining with Oil Red O for 5 min. After

staining, the plates were washed twice in water and

photographed. For the AdipoRed Assay, the medium was

removed in both untreated and DIM-induced rPADs, and

each well was carefully rinsed with 200 ml PBS. Then, the

rPADs were incubated with 200 ml PBS and 5 ml of

AdipoRed at RT for 10–15 min and immediately placed

in a fluorimeter for fluorescence measurement (excitation

at 485 nm and emission at 572 nm). Triglyceride content

was normalized on protein content. Both untreated and

DIM-treated rPADs, AdipoRed stained, were imaged

immediately using a Leica DMI6000 microscope equipped

with a DFC350FX camera. The images were acquired using

the Leica N3 filter set and a Fluotar 20! 0.4NA long-

working distance objective with a correction collar.

AdipoRed-positive cells, identified as those clearly exhibit-

ing lipid droplet staining, were counted using the ImageJ

software and expressed as the percentage of total cells.

Confocal microscopy

DIM-treated rPADs, following AdipoRed staining, were

immediately imaged using a Leica SP2-AOBS confocal

microscope, as described previously (Maneschi et al. 2012).

The images were collected as z-stacks through a 63! 1.2NA

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

water-immersion objective, taking care to minimize spheri-

cal aberration, and then deconvolved with the Huygens

Professional Software (Scientific Volume Imaging (SVI),

Hilversum, The Netherlands) using the Classic Maximum

Likelihood Estimation (CMLE) algorithm and a theoretical

Point Spread Function (PSF). Finally, these images were

quantitatively analyzed using the Volocity 5 Software

(Perkin-Elmer, Foster City, CA, USA) to measure the number

and volume of the lipid droplets.

Glucose uptake

Glucose uptake by rPADs was measured as described

previously (Maneschi et al. 2012). DIM-exposed rPADs

were cultured for 24 h in a serum-free medium, followed

by incubation in increasing concentrations of insulin

(1, 5, 10, and 50 nM) diluted in glucose-free Krebs

phosphate buffer (2.5 mmol Ca2C and 1 mg/ml BSA), to

evaluate insulin-dependent stimulation. At the end of the

incubation period, rPADs were further incubated with3H-2-deoxy-D-glucose (16 mM (1 mCi/ml); ICN Pharma-

ceuticals, Costa Mesa, CA, USA) for 5 min. The cells were

then washed with PBS and lysed with NaOH 0.5 M,

and the incorporated radioactivity was measured by

scintillation spectrometry using a b-counter (Perkin-Elmer).

Data were normalized on protein content.

RNA extraction and quantitative RT-PCR analysis

The isolation of RNA from the tissue and cells was

performed as described previously (Morelli et al. 2012,

2013). Specific primers for all the target genes have been

described previously (Filippi et al. 2009, Maneschi et al.

2012, Morelli et al. 2012, 2013, Vignozzi et al. 2012). The

expression of the 18S rRNA subunit was quantified with a

predeveloped assay (Applied Biosystems).

Statistical analysis

Results are expressed as meansGS.E.M. for n experiments as

specified. The statistical analysis was performed with a

one-way ANOVA test followed by the Tukey–Kramer

post hoc analysis in order to evaluate differences between

the groups, and P!0.05 was considered significant.

Correlations were assessed using Spearman’s method,

and the statistical analysis was performed with the

Statistical Package for the Social Sciences (SPSS, Inc.) for

Windows 15.0. Stepwise multiple linear regressions

were applied for the multivariate analysis, whenever

appropriate. Half-maximal response effective concentration

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 220

(EC50) values and maximal effect (Emax) values were

calculated using the computer program ALLFIT (De Lean

et al. 1978).

Results

OCA ameliorates HFD-induced metabolic alterations and

VAT dysfunction

Feeding rabbits the HFD enhanced all the components of

MetS, including insulin resistance and visceral fat

accumulation. The prevalence of three or more MetS

factors, identifying the MetS condition, according to the

human definition (Alberti et al. 2009), was verified in

62.9% of the HFD rabbits (P!0.001 vs CON). Table 2

Table 2 Metabolic and hormonal parameters in the experimen

evaluated using quantitative RT-PCR in VAT samples of the CON (nZ

groups. Data were calculated according to the comparative Ct

normalization

CON (nZ38)

Total body weight (g)Baseline 3258.6G63.0Week 12 3909.9G38.3‡

Blood glucose (g/l)Baseline 1.18G0.04Week 12 1.25G0.03OGTT (iAUC)Week 12 157.0G5.0Cholesterol (mg/dl)Baseline 36.9G2.1Week 12 42.8G3.0Triglycerides (mg/ml)Baseline 81.5G4.4Week 12 96.5G4.5AST (U/l)Baseline 30.9G3.1Week 12 35.9G3.0ALT (U/l)Baseline 25.7G2.2Week 12 28.5G1.9Liver weight (g, % of total body weight)Week 12 2.9G0.1MAP (mmHg)Week 12 91.5G2.2VAT weight (g, % of total body weight)Week 12 0.92G0.05Gallbladder (mg, % of total body weight)Week 12 27.6G4.617b-Estradiol (pmol/l)Week 12 168.7G91Presence of MetS (%) 0FXR expression in VAT (mRNA/18S) 55.62G9.21TGR5 expression in VAT (mRNA/18S) 11.42G1.32

iAUC, incremental area under the curve of glucose blood level during oral gluaminotransferase; MAP, mean arterial pressure; VAT, visceral adipose tissue. *P!vs CON week 12; and aP!0.05, bP!0.01, and cP!0.001 vs HFD week 12.

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

reports in detail the effects of the HFD on the metabolic

and hormonal parameters. Treatment with the selective

FXR agonist OCA for 12 weeks significantly reduced

fasting blood glucose levels and glucose intolerance

(both P!0.01 vs HFD). In addition, VAT weight was

markedly decreased, even after normalization for total

body weight (P!0.001 vs HFD). VAT amount in the

OCA-treated HFD rabbits was even below the CON level

(P!0.001; Table 2). In particular, OCA treatment was

able to significantly reduce the prevalence of MetS

(P!0.01 vs HFD, see Table 2). Conversely, treatment

with the specific TGR5 agonist INT-777 did not exert any

significant effect either on the prevalence of MetS (PZ0.09

vs HFD) or on glycemia and glucose intolerance. INT-777

induced only a significant reduction in the amount of VAT

tal rabbits. Relative mRNA expression of FXR and TGR5 was

38), HFD (nZ36), HFDCOCA (nZ18), and HFDCINT-777 (nZ6)

method using 18S rRNA subunit as the reference gene for

HFD (nZ36) HFDCOCA (nZ18) HFDCINT-777 (nZ6)

3290.4G42.0 3361.7G52.4 3266.6G107.53745.3G34.8‡ 3663.1G79.3* 3404G106.4s

1.29G0.03 1.25G0.06 1.24G0.21.94G0.07‡,¶ 1.40G0.06b 2.03G0.3*,s

224.8G7.4¶ 181.6G9.6b 217.1G15.7s

44.1G2.0 36.8G2.0 31.2G1.61447.6G64.7‡,¶ 1242.1G91.5‡,¶ 1711G134.6‡,¶

86.6G4.1 76.8G5.7 93.83G9.9304.5G25.1‡,¶ 230.7G36.6†,s 156.8G30.5

26.2G2.2 25.8G2.3 37.8G8.679.7G6.9‡,¶ 74.0G7.9‡,¶ 55.8G6.9§

23.0G1.6 27.3G3.1 34.3G4.646.1G3.1‡,¶ 35.6G4.3 33.3G6.3

4.24G0.1¶ 4.21G0.2¶ 3.72G0.2§

133.4G3.5¶ 129.2G4.3¶ 143.2G3.2¶

1.09G0.04s 0.41G0.06¶,c 0.51G0.09s,c

31.1G9.2 18.7G6 143.4G35.1¶,c

307.6G243¶ 135.5G50c 107G9.2§,c

62.9¶ 8.3b 100¶

60.99G12.76 97.28G6.3¶,b 66.69G15.529.97G1.92 11.20G1.89 17.25G0.87s,a

cose tolerance test (OGTT); AST, aspartate aminotransferase; ALT, alanine0.05, †P!0.01, and ‡P!0.001 vs baseline; §P!0.05, sP!0.01, and ¶P!0.001

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 221

(even after normalization for body weight; P!0.001 vs

HFD). In addition, a peculiar increase in gallbladder

weight was observed in the HFD rabbits treated with

INT-777 (P!0.001 vs HFD and P!0.001 vs CON). The

parameters of the relative transcripts of FXR and TGR5

in VAT isolated from the different experimental groups

are also reported for comparison. In the CON rabbits,

the parameters of the transcripts of FXR were fivefold

higher (P!0.001) than that of the transcripts of TGR5,

which were both unaffected by the HFD (see Table 2). OCA

treatment but not INT-777 treatment further increased

FXR mRNA expression (P!0.01 vs HFD). Conversely,

INT-777 significantly upregulated TGR5 gene expression

(P!0.05 vs HFD), but not FXR gene expression. Because no

significant changes were observed regarding HFD-induced

hyperglycemia and glucose intolerance in the INT-777-

treated HFD rabbits, no further studies were performed

on this particular group.

We next analyzed the correlation between the

expression of the FXR gene and that of several genes

related to inflammation, adipogenesis, glucose transport,

and insulin signaling in the VAT of rabbits fed the CON

or HFD. As shown in Table 3, a significant positive

relationship was found between the expression of the

FXR gene and that of the genes related to adipogenesis

(CCAAT enhancer binding protein-a (c/EBPa (CEBPA)),

peroxisome proliferator-activated receptor g (PPARg

(PPARG)), fatty acid binding protein 4 (FABP4), leptin,

adiponectin, PPARa (PPARA), and phospholipase A2

(PLPA2)), glucose transport (GLUT4 (SLC2A4), ras

Table 3 Association between FXR mRNA and VAT-specific

genes in VAT

r P value n

Adipogenesisc/EBPa 0.444 0.005 39PPARg 0.449 !0.0001 59FABP4 0.378 0.004 56Adiponectin 0.405 0.002 58Leptin 0.315 0.017 57PPARa 0.481 !0.0001 59PLPA2 0.359 0.012 48Glucose transport and insulin signalingGLUT4 0.347 0.007 60RHOA 0.280 0.030 60ROCK1 0.275 0.034 60ROCK2 0.375 0.003 60InflammationIL6 0.428 0.001 56MCP1 0.334 0.013 55

Correlations coefficients (r) and level of significance (P value) were derivedfrom the univariate analysis.

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

homolog gene family, member A (RHOA), Rho-associated,

coiled-coil-containing protein kinase 1 (ROCK1), and

ROCK2), and inflammation (interleukin 6 (IL6) and

monocyte chemoattractant protein-1 (MCP1 (CCL2))).

The histomorphometric analysis of adipocytes, and

the hypoxic state, along with perilipin expression, in the

three experimental groups are shown in Fig. 1. The HFD

induced a significant increase in adipocyte diameter

(P!0.01 vs CON; Fig. 1A, B and D), hypoxyprobe staining

(P!0.05 vs CON; Fig. 1E, F and H), and perilipin

expression (P!0.01 vs CON; Fig. 1I). All the parameters

were reduced by OCA treatment when compared with not

only the HFD (P!0.0001), but also the CON (P!0.0001)

(see bar graphs in Fig. 1D, H and I). In addition, GLUT4

expression in the adipocyte membrane fraction was

decreased and GLUT4 was stacked in the cytosol of the

HFD rabbits (P!0.01 vs CON), while OCA completely

normalized GLUT4 membrane translocation (P!0.05 vs

HFD; Fig. 1J).

The correlation of visceral fat weight and several of the

aforementioned VAT genes is reported in Table 4.

Essentially, increased visceral fat accumulation was

positively associated with the genes related to adipogen-

esis (c/EBPa, FABP4, and leptin), lipogenesis (diacyl-

glycerol O-acyltransferase (DGAT2) and lipoprotein

lipase (LPL)), NO signaling (endothelial nitric oxide

synthase (eNOS (NOS3)) and protein kinase G1 (PKG1)),

glucose transport (GLUT4, RHOA, ROCK1, ROCK2, and

vimentin (VIM)), inflammation (MCP1), steroid sensitivity

((estrogen receptor a (ERa (ESR1))), and cytoskeleton

remodeling (a smooth muscle actin (aSMA)). As shown

in Table 5, in vivo OCA dosing induced a downregulation

of the expression of most of these genes, including the

progesterone receptor, indicating a decreased estrogen

action (see Table 2). Conversely, OCA dosing upregulated

the expression of the FXR downstream gene small

heterodimer partner (SHP, Table 5).

OCA ameliorates HFD-induced liver steatosis

and inflammation

As reported in Table 2, the HFD also induced a significant

increase in liver weight, as well as aspartate aminotransfer-

ase serum levels, which were not normalized by OCA

dosing. We then evaluated FXR and TGR5 relative mRNA

expression in the liver. We found that TGR5 was 2-log unit

less expressed than FXR (29.51G6.6 and 6177.92G851.16

respectively). The correlation between the expression of

liver FXR and that of several genes related to steatosis,

inflammation, fibrosis, and metabolism in the CON and

Published by Bioscientifica Ltd.

A CON B HFD C HFD+OCA

E CON F HFD G HFD+OCA

D

30

CONHFD

HFD+OCA

CONHFD

HFD+OCA

60

90

Adi

pocy

te d

iam

eter

(µm

)

**

***°°°

30

60

90

120

150

Hyp

oxyp

robe

pos

itivi

ty(%

of C

ON

)

H

*

***°°°

I

**

***°°°

Perilipin

Actin

J

mGLUT4

cGLUT4

0

100

200

300

0

50

100

150

HFD HFD+OCA

mG

LUT

4/cG

LUT

4

Per

lipin

OD

(%

of C

ON

)

**

°

CONHFD HFD+OCACON

Figure 1

Effects of OCA treatment on the morphological and functional features of

VAT in the experimental rabbits. (A, B and C) Representative images of the

hematoxylin and eosin-stained VAT sections showing different adipocyte

sizes among the experimental groups (magnification 20!, scale barZ

50 mm). Adipocyte size was significantly increased in the HFD rabbits

when compared with that in the CON and OCA-treated HFD rabbits.

(D) Histomorphometric analysis of adipocyte diameter (mm) in the different

experimental groups (nZ3 for each group). (E, F and G) Immunohisto-

chemical staining of hypoxyprobe adducts in VAT sections. Hypoxyprobe

adducts were revealed in hypoxic cells (PO2 !10 mmHg) of VAT transverse

sections by a MAB (magnification 10!, scale barZ50 mm). An intense

hypoxyprobe positivity was detected in VAT isolated from the HFD rabbits

(F), while only scanty positive labeling was present in VAT isolated from the

CON (E) and OCA-treated HFD (G) rabbits. (H) Computer-assisted

quantitative image analysis of three independent experiments (nZ3 for

each group). (I) Protein expression of perilipin in VATextracts isolated from

the experimental rabbits. Representative immunoblots with anti-perilipin

and anti-actin primary antibodies and the corresponding graphical

representation of optical density (OD) analysis of perilipin band intensity

normalized over actin are shown (nZ5 for each group). (J) Analysis of

GLUT4 membrane translocation in VAT. The lower panel shows represen-

tative immunoblots with anti-GLUT4 primary antibody on the membrane

(m) and cytosolic (c) fractions of VAT isolated from the CON, HFD, and

OCA-treated HFD rabbits. The bar graph shows the optical density analysis

of membrane:cytosolic GLUT4 ratio (nZ5 for each group). Data are

expressed as the percentage of CON values. *P!0.05, **P!0.01, and

***P!0.0001 vs CON; 8P!0.05 and 888P!0.0001 vs HFD. Full colour version

of this figure available via http://dx.doi.org/10.1530/JOE-13-0109.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 222

HFD rabbits is reported in Table 6. As shown in Table 7, the

mRNA expression of FXR and the FXR primary response

gene cholesterol 7a-hydroxylase (CYP7A1) was signi-

ficantly increased in the liver of the HFD rabbits (P!0.05

and P!0.01 respectively). OCA dosing upregulated the

expression of the FXR (P!0.01 vs CON) and SHP (P!0.001

vs CON and P!0.01 vs HFD) genes, while the expression

of the CYP7A1 gene was downregulated (P!0.05 vs CON

and P!0.0001 vs HFD; see also Table 7). Immunohisto-

chemical studies using Oil Red O staining revealed a

homogeneous and abundant hepatic lipid deposition in

the HFD rabbits when compared with the CON rabbits

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

(P!0.0001; Fig. 2A and B). OCA dosing was able to

markedly counteract lipid accumulation, which was

mainly limited to the perilobular region occupied by the

portal system (Fig. 2C). The quantitative computer-

assisted analysis of Oil Red O staining is shown in

Fig. 2D. Gene expression of PPARg, a specific steatosis

marker, was significantly increased in the HFD rabbits

(P!0.001 vs CON) and normalized by OCA dosing

(P!0.01 vs HFD; Fig. 2E). Similar results were obtained

for adiponectin mRNA (P!0.01 vs CON and P!0.01 vs

HFD; data not shown). The livers isolated from the HFD

rabbits also exhibited an intense intrahepatocyte

Published by Bioscientifica Ltd.

Table 4 Association between visceral fat weight and VAT-

specific genes in VAT

r P value n

Adipogenesisc/EBPa 0.296 0.05 52FABP4 0.392 !0.0001 70Leptin 0.396 !0.0001 70LipogenesisDGAT2 0.379 0.001 68LPL 0.240 0.05 71NO signalingENOS 0.291 0.01 74PKG1 0.249 0.05 74Glucose transportGLUT4 0.378 !0.0001 74RHOA 0.373 !0.0001 74ROCK1 0.325 0.01 72ROCK2 0.289 0.01 71VIM 0.258 0.05 68InflammationMCP1 0.374 0.001 74Steroid sensitivityERa 0.362 0.01 57Cytoskeleton remodelingaSMA 0.316 0.05 61

Correlations coefficients (r) and level of significance (P value) were derivedfrom the univariate analysis.

Table 5 Effect of OCA treatment on the mRNA expression of

VAT-specific genes. Data are expressed as the percentage of

variation vs HFD

Genes

Percentage of variation

(HFDCOCA vs HFD)

SHP 274.3G92.6†

FABP4 K47G11.3†

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 223

immunopositivity for anti-TNFa antibody (P!0.01 vs

CON; Fig. 2F and G), which was significantly blunted by

OCA dosing (P!0.01 vs HFD; Fig. 2H). The quantitative

computer-assisted analysis of anti-TNFa staining is shown

in Fig. 2I. The expression of inflammation genes, TNFa

(P!0.001; Fig. 2J), IL6 (data not shown; P!0.05), and IL10

(data not shown; P!0.001), was significantly increased in

the liver of the HFD rabbits when compared with the CON

rabbits. OCA dosing normalized the expression of both

TNFa (Fig. 2J) and IL6 (data not shown), while it

significantly increased that of IL10 (P!0.05, data not

shown).

c/EBPa K61.2G12.3‡

LPL K49.6G7.9*Leptin K58.2G23*GLUT4 K31.7G8.7*IRS1 K32G3.9†

RHOA K37G8.2†

ROCK1 K34.8G7.8†

ROCK2 K56G16.1†

DGAT2 K63.5G17.3*PR (PGR) K42.3G8.1*VIM K17.7G2.3aSMA K48.8G15.8MCP1 K13.7G5.1eNOS K4.8G1ERa K22G5.8PKG1 K21.4G4.7

*P!0.05, †P!0.01, and ‡P!0.001 vs HFD.

OCA ameliorates spontaneous adipogenic differentiation

in rabbit preadipocytes

We next investigated the adipogenic capacity of rPADs

isolated from VAT. Each cell preparation was characterized

by flow cytometry for the expression of mesenchymal

stem cell (MSC) markers and hematopoietic–monocytic

contamination. The percentage of positive cells expressing

the MSC markers CD90, CD105, and CD106 was not

different among the groups (data not shown). All rPADs

were negative for endothelial (CD31), hematopoietic

(CD34 and CD45), and monocytic (CD14) markers (data

not shown). Expression analysis by qRT-PCR showed

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

that the expression of CD90 (THY1) and the specific

marker of adipocyte commitment dickkopf 1 (DKK1)

was not different among the groups (data not

shown). Interestingly, preadipocytes isolated from the

OCA-treated rabbits exhibited a significantly increased

expression of adipogenic-specific genes, such as FABP4

(P!0.001), c/EBPa (P!0.01), and PPARg (P!0.01),

compared with those isolated from both the CON and

HFD groups (data not shown).

The spontaneous adipogenic potential was investi-

gated in rPADs cultured for 10 days (Fig. 3A, B and C). The

qualitative (Oil Red O staining; Fig. 3A) and quantitative

(AdipoRed assay; Fig. 3B) estimation of triglyceride

accumulation showed an increased lipid content in the

cytosol of rPADs isolated from the OCA-treated HFD

rabbits when compared with those isolated from both

the HFD and CON groups. The HFD-reduced percentage

of the AdipoRed-positive cells was also completely

normalized by OCA dosing (P!0.01), even significantly

higher than that in the CON group (P!0.01; Fig. 3C).

OCA ameliorates DIM-induced adipogenic

differentiation in rPADs

We next evaluated the adipogenic potential by exposing

in vitro rPADs to a DIM for 10 days (Fig. 3D, E and F).

Oil Red O staining and AdipoRed assay showed a reduced

Published by Bioscientifica Ltd.

Table 6 Association between the expression of FXRmRNA and

that of other genes related to steatosis, metabolism, inflam-

mation, and fibrosis in the liver

r P value n

SteatosisPPARg 0.623 !0.0001 69Adiponectin 0.324 0.007 67MetabolismPPARa 0.419 !0.0001 69PLPA2 0.429 0.006 52InflammationTNFa 0.377 0.002 67IL6 0.291 0.017 67MCP1 0.382 0.001 69COX2 (PTGS2) 0.388 0.001 68IL8 0.509 !0.0001 61IL10 0.455 !0.0001 61CD4 0.248 0.046 65CD8 0.395 0.003 53CD68 0.445 !0.0001 65FibrosisaSMA 0.563 !0.0001 62RHOA 0.636 !0.0001 66ROCK1 0.569 !0.0001 66ROCK2 0.421 0.001 64TGFb (TGFB1) 0.496 !0.0001 63COL1A1 0.412 0.002 53COL3A1 0.505 !0.0001 53TIMP1 0.530 !0.0001 63TIMP2 0.672 !0.0001 50MMP2 0.641 !0.0001 50MMP9 0.551 !0.0001 49

Correlation coefficients (r) and level of significance (P value) were derivedfrom the univariate analysis.

Table 7 Expression of genes involved in FXR activation in livers

isolated from all the rabbit groups. Expression of genes

involved in FXR activation (FXR, SHP, and CYP7A1) was detected

by qRT-PCR in livers isolated from all the rabbit groups

FXR SHP CYP7A1

CON (nZ31) 100G6.6 100G12.05 100G14.35HFD (nZ36) 131.4G11* 188.4G42.1 300.4G75†

HFDCOCA(nZ18)

145.1G10.6† 336.8G58.7‡,§ 81G34.7*,s

Data are expressed as the percentage of CON *P!0.05, †P!0.01, and‡P!0.0001 vs CON; §P!0.01 and sP!0.0001 vs HFD.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 224

adipogenic differentiation, characterized by a reduced

triglyceride content (Fig. 3E) and a reduced percentage of

AdipoRed-positive cells (Fig. 3F), in rPADs isolated from

the HFD rabbits when compared with those isolated from

the CON rabbits (both P!0.01). OCA treatment of the

HFD rabbits completely normalized the percentage of

AdipoRed-positive cells (Fig. 3F) and triglyceride content

(both P!0.01 vs HFD), with the latter being even higher

than that in the CON rabbits (P!0.01; Fig. 3D and E).

The responsiveness of rPADs to the DIM was also

investigated in terms of the expression of adipocyte-

related genes (DKK1, c/EBPa, PPARg, FABP4, adiponectin,

and leptin). As reported in Table 8, after 10 days of

exposure to the DIM, there was a significant induction of

the expression of all the investigated genes in rPADs

isolated from the CON rabbits (all genes P!0.01 vs relative

time 0). Conversely, in rPADs isolated from the HFD

rabbits, DIM exposure was unable to significantly induce

the expression of the investigated genes, with the

exception of FABP4 mRNA. OCA treatment normalized

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

the DIM-induced expression of all these adipocyte-specific

genes (Table 8). Similarly, cyclin D3 (CCND3) mRNA

expression was significantly induced in all the DIM-

treated rPADs, with the exception of rPADs isolated from

the HFD rabbits (Table 8). Conversely, CCND1 mRNA

expression was significantly increased only in DIM-treated

rPADs isolated from the HFD group and not in those

isolated from the other groups (P!0.05, Table 8).

Lipid droplets in rPADs isolated from the HFD rabbits

exhibited a reduction in the average number and an

increase in the average volume per cell when compared

with those in rPADs isolated from the CON rabbits (P!0.05

andP!0.0001 respectively; Fig. 4A, B, D and E). In vivoOCA

dosing induced both an increase in the number (P!0.0001;

Fig. 4C and D) and a reduction in the volume (P!0.0001;

Fig. 4C and E) of lipid droplets when compared with

those in rPADs isolated from the HFD rabbits.

Using qRT-PCR, we observed a significant upregulation

of the expression of genes of the SNARE complex involved in

lipid droplet handling, synaptosomal-associated protein 23

(SNAP23) and syntaxin 5 (SYNT5), in both untreated

and DIM-induced rPADs isolated from the HFD rabbits

when compared with those isolated from the CON rabbits

(Table 9). In vivo OCA dosing normalized the expression of

these genes (Table 9). The expression of SNAP23 and SYNT5

in both untreated and DIM-induced rPADs isolated from all

the groups, expressed as a function of lipid droplet volume,

is shown in Fig. 4F and G. A significant positive relationship

was found between the lipid droplet volume and SNAP23

(rZ0.928, PZ0.008; Fig. 4F) and SYNT5 (rZ0.829, PZ0.04;

Fig. 4G). Conversely, in vivo OCA dosing had effects

comparable to those of the CON.

OCA ameliorates glucose uptake in rPADs

The effect of OCA on insulin sensitivity was investigated

by measuring 3H-2-deoxy-D-glucose uptake in DIM-induced

Published by Bioscientifica Ltd.

A CON B HFD C HFD+OCA

0

10

CONHFD

HFD+OCA

CONHFD

HFD+OCA

CONHFD

HFD+OCA

CONHFD

HFD+OCA

20

30

40

50

Oil

red

O p

ositi

vity

(O

D)

***

***

D

°°°

F CON G HFD H HFD+OCA

0

5

10

15

20

25

TN

Fα

posi

tivity

(O

D)

I

°°

*

0

100

200

300

400

500

PP

ARγ

mR

NA **

°°

0

100

200

300

400

TN

Fα

mR

NA

**

°

E

J

Figure 2

Amelioration of HFD-induced liver steatosis and inflammation byOCA. (A, B

and C) Lipid accumulationwas revealed in liver sections of the experimental

rabbits by Oil Red O staining (magnification 10!, scale barZ50 mm). An

abundant hepatic lipid deposition was found in the HFD rabbits (B) when

compared with the CON rabbits (A). OCA dosing was able to markedly

counteract lipid accumulation, mainly limited to the perilobular region

occupied by the portal system (C). The quantitative computer-assisted

analysis ofOil RedO staining is shown in (D). (E) RelativemRNAexpressionof

steatosis marker (PPARg) was evaluated using quantitative RT-PCR in liver

samples of the CON (nZ38), HFD (nZ36), and HFDCOCA (nZ18) rabbits.

Datawere calculated according to the comparative Ct method using the 18S

rRNA subunit as the reference gene for normalization. Results are expressed

as percentage over the CON. (F, G and H) Immunohistochemistry for TNFa in

liver sections of the experimental rabbits (magnification 20!, scale barZ

50 mm). Livers isolated from the HFD rabbits (G) exhibited an intense

intrahepatocyte immunopositivity for anti-TNFa antibody, when compared

with those isolated fromtheCONrabbits (F),whichwas significantlyblunted

byOCAdosing (H). The quantitative computer-assisted analysis of anti-TNFa

staining is shown in (I). (J) RelativemRNAexpressionof inflammationmarker

(TNFa) was evaluated using quantitative RT-PCR in the liver samples of the

CON (nZ38), HFD (nZ36), and HFDCOCA (nZ18) rabbits. Data were

calculated according to the comparative Ct method using 18S rRNA subunit

as the referencegene fornormalization. Results are expressedaspercentage

over the CON. *P!0.01, **P!0.001, and *** P!0.0001 vs CON; 8P!0.05,

88P!0.01, and 888P!0.0001 vsHFD. Full colour version of this figure available

via http://dx.doi.org/10.1530/JOE-13-0109.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 225

rPADs, after exposure to increasing concentrations of

insulin. As shown in Fig. 5, insulin dose dependently

stimulated glucose uptake in rPADs isolated from the three

experimental groups with significant differences for both

EC50 and Emax (P!0.0001). In vivo OCA dosing restored the

normal sensitivity to insulin (CON and HFDCOCA shared

EC50Z2.96G0.51 nM; HFD EC50Z13.5G6.09 nM). The

Emax of the HFD rabbits was dramatically decreased

(128G4%) when compared with that of both the CON

(CON EmaxZ273G3%, PZ0.001) and OCA-treated HFD

(HFDCOCA EmaxZ205G3%, PZ0.004) groups, although

the Emax of the latter group was still lower than that of

the CON rabbits (PZ0.006).

Discussion

In this study, we demonstrate that pharmacological

activation of FXR by OCA treatment prevents several

HFD-induced alterations in the liver, while normalizing

hyperglycemia and glucose intolerance as well as all the

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

MetS-related VAT dysfunctions, including preadipocyte

differentiation toward a mature phenotype and lipid

droplet handling.

This study, in addition to confirming previous results

(Maneschi et al. 2012), highlights several novel aspects in

the relationship between AT and MetS. Our studies were

carried out using a non-genomic, rabbit model of MetS,

which essentially recapitulates the human phenotype

(Filippi et al. 2009, Vignozzi et al. 2011, 2012, Maneschi

et al. 2012, Morelli et al. 2012, 2013). Feeding a HFD for

12 weeks induces a sharp increase in fasting glycemia,

glucose intolerance, and VAT amounts, as well as

hypertension and dyslipidemia. In the HFD-induced rabbit

model of MetS, VAT is not only increased in mass but also

dysfunctional, with an impaired triglyceride synthesis and

insulin-stimulated adipogenesis. We demonstrate that

this animal model of MetS is also characterized by liver

inflammation and steatosis, the main features of NASH.

There is a close relationship between VAT dysfunction and

NASH in MetS. Insulin resistance is the putative key

Published by Bioscientifica Ltd.

CA

0

400

800

1200

1600

Lipi

d co

nten

t/µg

prot

ein

(RF

U)

^ ^

40

50

60

70

Per

cent

age

of A

dipo

Red

posi

tive

cells

*

*°

B

D

40

50

60

70

80

90

Per

cent

age

of A

dipo

Red

posi

tive

cells

^

E

0

1000

2000

3000

4000

HFD

HFD+OCA

Lipi

d co

nten

t/µg

prot

ein

(RF

U)

*

*°

CONHFD

HFD+OCA

CON

HFD

HFD+OCA

CONHFD

HFD+OCA

CON

F

HFD HFD+OCA

CON HFD HFD+OCA

CON

Figure 3

Amelioration of both spontaneous and DIM-induced adipogenic differen-

tiation in rPADs by in vivo OCA dosing. (A) Lipid content (white arrows) in

untreated rPADs isolated from each experimental group, as evaluated by

Oil Red O staining. (B) Quantitative assessment of lipid content in

untreated rPADs isolated from each experimental group, as evaluated by

the AdipoRed assay. Results are expressed as relative fluorescence unit

(RFU)/mg of protein (nZ6 for each group). (C) Analysis of the lipid droplet

content in untreated, AdipoRed-stained, rPADs isolated from all the rabbit

groups. AdipoRed-positive cells were counted using the ImageJ Software

and are expressed as the percentage of total cells. (D) Lipid content in DIM-

induced rPADs isolated from each experimental group, as evaluated by Oil

Red O staining. (E) Quantitative assessment of lipid content in DIM-induced

rPADs isolated from each experimental group, as evaluated by the

AdipoRed assay. Results are expressed as RFU/mg of protein (nZ6 for each

group). (F) Analysis of the lipid droplet content in DIM-exposed, AdipoRed-

stained, rPADs isolated from all the rabbit groups. AdipoRed-positive cells

were counted using the ImageJ Software and are expressed as the

percentage of total cells. ^P!0.01 and ^^P!0.001 vs all the other groups;

*P!0.01 vs CON; and 8P!0.01 vs HFD. Full colour version of this figure

available via http://dx.doi.org/10.1530/JOE-13-0109

Table 8 Effect of DIM on the mRNA expression of adipocyte-

related genes in rPADs. Relative mRNA expression of adipocyte-

related genes was evaluated using quantitative RT-PCR in

untreated (time 0) and DIM-exposed rPADs from the CON,

HFD, and HFDCOCA groups (five different experiments, each

performed in triplicate using a different cell preparation per

group). Data were calculated according to the comparative Ct

method using 18S rRNA subunit as the reference gene for

normalization. Results are expressed as fold change over time 0

CON HFD HFDCOCA

Adipocyte-related genesDKK1 6.4G2† 1.5G0.3‡ 13.6G2.1†

c/EBPa 2.3G0.5† 1.2G0.3‡ 2.5G0.5†

PPARg 2.5G0.5† 1.1G0.3‡ 1.7G0.1†

FABP4 20.6G7† 5.3G1.1† 10.9G3.6†

Adiponectin 9.5G4.3† 0.9G0.1‡ 2.6G0.7†

Leptin 8.7G2.6†,‡ 0.7G0.2 1.8G0.4CCND1 0.8G0.3 2.6G0.7* 1.1G0.1CCND3 2.3G0.5*,§ 1.2G0.3 1.9G0.3†,s

*P!0.05 and †P!0.01 vs relative time 0; ‡P!0.01 vs all the other groups;and §P!0.05 and sP!0.01 vs relative CCND1.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 226

underlying mechanism linking these two clinical entities

(Cusi 2012, Targher & Byrne 2013).

In the present model of MetS, HFD induced a

significant increase in liver weight and an abundant lipid

accumulation, which were associated with an increased

expression of steatosis markers, such as PPARg and

adiponectin. Livers isolated from MetS rabbits were also

severely inflamed, as demonstrated by an increased

expression of TNFa and IL-6 – pro-inflammatory cytokines

involved in the transition from NAFLD to NASH. Indeed,

the activation of inflammatory pathways in NASH is

related to hepatic toxicity resulting from intrahepatic

triglyceride overload (Cusi 2012). The major contributor

to an increased triglyceride accumulation is dysfunctional

AT (Donnelly et al. 2005). Interestingly, as described

previously (Maneschi et al. 2012), we confirmed that

VAT adipocytes isolated from MetS animals are dysfunc-

tional. An increase in size and hypoxia, along with a

reduced membrane translocation of GLUT4 and an

increased expression of perilipin, was observed in VAT

adipocytes isolated from the HFD rabbits. Indeed, not only

the total mass of AT conveys a metabolic risk, but the size

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

Published by Bioscientifica Ltd.

A CON B HFD C HFD+OCA

D E

1 2 3 4

Lipi

d dr

ople

ts v

olum

e (µ

m3 )

SNAP23 mRNA

F

HFD DIM

HFD untreated

HFD+OCA DIM

CON DIM

CON untreated

HFD+OCA untreated

1 6 11 16 21

Lipi

d dr

ople

ts v

olum

e (µ

m3 )

SYNT5 mRNA

G

rPAD:

0

200

400

600

HFD HFD+OCA

Mea

n nu

mbe

r of

lipi

ddr

ople

ts

*

**°°°

CON0.0

0.5

1.0

1.5

2.0

2.5

HFD HFD+OCA

Mea

n vo

lum

e (µ

m3 )

of

lipid

dro

plet

s

***

*°°°

CON

HFD DIM

HFD untreated

HFD+OCA DIM

CON DIM

CON untreated

HFD+OCA untreated

rPAD:

2.5

2.0

1.5

1.0

0.5

0.0

2.5

2.0

1.5

1.0

0.5

0.0

Figure 4

Positive effect of OCA on lipid droplet fusion. rPADs isolated from the CON

(A), HFD (B), and OCA-treated HFD (C) rabbits were imaged by confocal

microscopy (scale barZ10 mm). Images were quantitatively analyzed using

the Volocity 5 Software (Perkin-Elmer, Foster City, CA, USA) to measure the

number (D) and volume (mm3; E) of lipid droplets within single cells. At least

eight cells were analyzed for each group. (F and G) Relationship between

the lipid droplet volume (expressed as mm3, ordinate) and the SNAP23 or

SYNT5 mRNA expression (abscissa) in both untreated and DIM-induced

rPADs as derived from univariate Spearman’s regression analysis. *P!0.05,

**P!0.01, and ***P!0.0001 vs CON; 888P!0.0001 vs HFD.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 227

of adipocytes is also important, being positively associated

with insulin resistance (Jacobsson & Smith 1972, Salans

et al. 1974). Findings regarding the normalization of

insulin resistance after weight loss, associated with a

reduction in adipose cell size (Salans et al. 1968), further

corroborate this concept. Interestingly, a putative

mechanism by which insulin resistance could develop in

hypertrophic fat cells may originate from hypoxia.

Previous studies have indicated that hypoxia develops in

VAT, as adipocyte size and tissue mass increase, leading to

– via different mechanisms, including reduction in the

expression of GLUT4 – an insulin-resistant phenotype

(O’Rourke et al. 2011, Trayhurn 2013). Accordingly, a

reduced GLUT4 translocation to the plasma membrane

was observed in hypertrophic fat cells, when compared

with the smaller ones (Salans et al. 1968, Salans &

Dougherty 1971, Smith 1971, Jacobsson & Smith 1972,

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

Olefsky 1976, Franck et al. 2007, Goossens 2007). More-

over, enrichment of perilipin 1 in large vs small adipocytes

has also been associated with reduced insulin sensitivity in

hypertrophic fat cells (Laurencikiene et al. 2011). Perilipin,

a reliable marker of adipogenesis, is a major anti-lipolytic

protein, coating the cytosolic surface of intracellular lipid

droplets, protecting or exposing the triacylglycerol core of

the droplets to lipases (Brasaemle 2007), thus controlling

access to the adipocyte triglyceride stores that supply

energy to most tissues. As a regulator of lipid storage and

lipolysis, perilipin 1 is thus positioned to modify not only

the risk of obesity but also its complications (Smith &

Ordovas 2012).

In the present study, we extensively investigated

insulin sensitivity and lipid droplet remodeling in

adipocytes. rPADs isolated from VAT of the HFD rabbits

exhibited a lower capacity to respond to insulin in terms

Published by Bioscientifica Ltd.

Table 9 Effect of in vivo OCA on the mRNA expression of

genes involved in lipid droplet fusion (SNAP23 and SYNT5).

Relative mRNA expression of the genes of the SNARE complex

involved in lipid droplet handling (SNAP23 and SYNT5) was

evaluated using quantitative RT-PCR in untreated (time 0) and

DIM-exposed rPADs isolated from the CON, HFD, and HFDC

OCA rabbits (six different rPAD preparations from each

experimental group)

SNAP23 SYNT5

CONUntreated 1.90G0.20 4.22G0.62DIM 2.09G0.23 3.88G0.36HFDUntreated 2.46G0.21* 9.90G2.44†

DIM 2.95G0.44* 12.02G3.33†

HFDCOCAUntreated 1.70G0.13s 3.52G0.39s

DIM 2.09G0.19§ 4.38G1.09§

*P!0.05, †P!0.01 vs CON; §P!0.05 and sP!0.01 vs HFD.

10–9 10–8 10–7

100

125

150

175

200

225

250

275

300

0

rPAD HFDrPAD CON

rPAD HFD + OCA

Insulin (M)

3 H-g

luco

se u

ptak

e (%

)

**°

*

Figure 5

Insulin sensitivity of DIM-exposed rPADs. Dose–response curves of

radiolabeled 3H-glucose uptake in DIM-treated rPADs after exposure to

increasing concentrations of insulin are shown. Results are expressed as

percentage over 0 nM insulin (five different experiments, each performed

in duplicate and using a different cell preparation per group). The relative

EC50s and Emax values are reported in the text. *P!0.01 and **P!0.001 vs

CON; 8P!0.01 vs HFDCOCA.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 228

of triglyceride synthesis and glucose uptake. Insulin

resistance in rPADs was also demonstrated by the failure

to upregulate the expression of adipogenesis-specific genes

such as DKK1, c/EBPa, PPARg, FABP4, adiponectin, and

leptin. In addition, DIM-exposed rPADs from VAT of the

HFD rabbits exhibited a prevalent expression of CCND1

when compared with the expression of CCND3 (Fu et al.

2004, Sarruf et al. 2005). Cyclins function as key

components of the cell-cycle core machinery in adipo-

cytes. Indeed it has been reported that CCND1 inhibits

adipocyte differentiation through the repression of the

expression and transactivation of PPARg, while CCND3

promotes adipocyte differentiation as the coactivator of

PPARg. Accordingly, a lower percentage of AdipoRed-

positive cells was also observed in the HFD rabbits.

These findings thus further support the view of

impaired adipocyte maturation in VAT isolated from the

HFD rabbits.

In addition, lipid droplets of rPADs isolated from the

MetS rabbits were reduced in number and increased in

volume, with an increased expression of factors involved

in lipid droplet fusion, namely the SNARE complex. Lipid

droplets are formed as primordial droplets and increase in

volume by a fusion process that requires the SNARE

complex, including SNAP23 and SYNT5 (Bostrom et al.

2007). Accordingly, in the present study, we found that

the expression of both SNAP23 and SYNT5 was increased

in rPADs isolated from the HFD rabbits. In addition, in

both untreated and DIM-induced rPADs, we found a

positive association between lipid droplet volume and

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

SNAP23 or SYNT5 mRNA expression. SNAP23 is also

required for insulin-stimulated translocation of GLUT4

to the plasma membrane (Foster et al. 1999, Kawanishi

et al. 2000), and it may play a role in the development

of insulin resistance. Indeed, when SNAP23 is diverted

from the plasma membrane, and thus away from the

mechanism involved in insulin-stimulated GLUT4

translocation and glucose uptake, it is instrumental in

the processes of lipid droplet fusion. This could represent

a putative mechanism by which the development of

insulin resistance is associated with the enhanced fusion

of lipid droplets.

The most striking feature of the present study is that

OCA treatment restores the differentiation of MetS

preadipocytes toward a more mature and efficient meta-

bolic phenotype, documented by their higher content of

small-volume lipid droplets, associated with a decreased

expression of factors known to orchestrate their fusion,

such as the SNARE complex, including SNAP23. Consist-

ent with the positive effect of OCA on HFD-induced

VAT dysfunction, DIM-exposed rPADs isolated from the

OCA-treated MetS rabbits exhibited an increased ability to

respond to insulin, in terms of glucose uptake and

adipocyte differentiation capacity, when compared with

rPADs isolated from the HFD rabbits. In addition, in

rPADs isolated from the OCA-treated HFD rabbits, all the

other DIM-induced adipocyte features, including tri-

glyceride synthesis, adipogenesis-specific gene expression

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 229

(DKK1, c/EBPa, PPARg, FABP4, adiponectin, and leptin),

preadipocyte maturation (CCND1 and CCND3), and the

number of differentiating cells (AdipoRed staining), were

also normalized. Interestingly, OCA exerts its pro-adipo-

genic effects even in the earlier stage of adipocyte

differentiation, as DIM-untreated preadipocytes isolated

from the OCA-treated HFD rabbits exhibited an increased

expression of adipogenesis-specific genes (such as c/EBPa,

PPARg, and FABP4, even when compared with those in the

CON rabbits) as well as a significant increase in both

triglyceride accumulation and percentage of differentiat-

ing cells. Overall, these findings are in line with previous

observations showing the effect of OCA on the promotion

of insulin sensitivity and adipocyte differentiation, both

in vivo (Cariou et al. 2006, Ma et al. 2006, Zhang et al. 2006)

and in vitro, as in the preadipocyte cell line 3T3-L1 (Cariou

et al. 2006, Rizzo et al. 2006). In the present study, we

demonstrated that this VAT weight reduction is associated

with adipocytes that are smaller in size. In vitro, we found

that preadipocytes isolated from the OCA-treated HFD

rabbits were able to differentiate into adipocytes with

multilocular lipid droplets and increased insulin sensi-

tivity. Interestingly, these phenotypic features have been

recognized to characterize the metabolically healthy

adipocytes, with increased energy consumption through

free fatty acid oxidation and consequently reduced fat

mass and insulin resistance (Timmons et al. 2007). An

increased free fatty acid oxidation could be the underlying

mechanism of the reduced visceral fat mass observed in

the OCA-treated HFD rabbits. A major limitation of the

present study is the lack of investigation on energy

consumption in preadipocytes isolated from the different

experimental groups. However, several recent studies have

demonstrated that the activation of FXR enhances energy

expenditure, reducing circulating levels of free fatty acids

and insulin resistance (Fiorucci et al. 2010).

The present study also indicates that FXR could be a

target for treating MetS-induced VAT alterations. In hom-

ogenates ofvisceral fat, indeed,we found that theexpression

of FXR is positively associated with the expression of genes

involved in insulin signaling and glucose transport (GLUT4,

RHOA, ROCK1, and ROCK2), adipogenesis (c/EBPa, PPARg,

FABP4, adiponectin, leptin, PPARa, and PLPA2), and

inflammation (IL6 and MCP1). Moreover, OCA dosing

completely normalized GLUT4 membrane translocation

and VAT oxygenation, as well as perilipin expression, and

drastically reduced adipocyte size, which was significantly

reduced even when compared with that observed in the

CON rabbits. OCA dosing also reduced the expression of

several genes associated with visceral fat accumulation,

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

including those related to inflammation (MCP1), steroid

sensitivity (ERa), adipogenesis (c/EBPa, FABP4, and leptin),

lipogenesis (DGAT2andLPL),NO signaling (eNOSandPKG),

glucose transport (GLUT4,RHOA,ROCK1,ROCK2, andVIM),

and cytoskeleton remodeling (aSMA).

Concomitantly, OCA also ameliorates HFD-induced

glucose intolerance and fasting hyperglycemia. The

increased insulin sensitivity may be responsible for the

preservation of ‘metabolically healthy’ VAT phenotype as

well as the amelioration of liver abnormalities. The present

data, showing that OCA dosing can reduce HFD-induced

liver steatosis and inflammation, as well as ALT serum

levels, are in line with previous results obtained in insulin-

resistant Zucker fa/fa rat model (Cipriani et al. 2010). The

reduced hepatic lipid levels correlate with the increased

insulin sensitivity in adipocytes, as reported previously

(Renga et al. 2010). Interestingly, OCA has been evaluated

in three phase II clinical trials, including one in patients

with type 2 diabetes and NAFLD (Adorini et al. 2012,

Mudaliar et al. 2013). In this trial, OCA was demonstrated to

induce a systemic improvement of insulin sensitivity and

an improvement in both hepatic and peripheral glucose

uptake. Interestingly, a significant decrease in the levels

of liver fibrosis biomarkers was also observed following

OCA treatment (Adorini et al. 2012, Mudaliar et al. 2013).

Our results indicate that the beneficial effect of OCA

on HFD-induced insulin resistance is mediated by the

specific activation of FXR, rather than TGR5, at both VAT

and hepatic levels. Indeed, we found that i) the treatment

of the HFD rabbits with the selective TGR5 agonist

INT-777 does not affect HFD-induced glucose intolerance

and increased fasting glycemia; ii) the expression of TGR5

in the liver and VAT is markedly lower compared to FXR;

iii) the expression of FXR primary response genes, SHP and

CYP7A1, is respectively upregulated and downregulated

by OCA treatment, as expected following FXR activation

(Rizzo et al. 2006). These data, together with the known

200-fold greater agonistic activity of OCA for FXR when

compared with TGR5 (Rizzo et al. 2010), support the view

that all the observed OCA effects on HFD-induced MetS are

selectively mediated by FXR activation.

In conclusion, in an animal model of HFD-induced

MetS, OCA dosing not only ameliorates liver steatosis and

inflammation but also counteracts all the HFD-induced

VAT alterations, restoring preadipocyte differentiation

through a positive and persistent effect on insulin

sensitivity. The present preclinical evidence and the

clinical experience with this first-in-class FXR agonist

support the potential of OCA to counteract diet-related

metabolic disorders.

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 230

Declaration of interest

E M, LV, A M, T M, S F, I C, P C, E S, A C, B M, R V, and G B V have nothing to

declare. L A is an employee of Intercept Pharmaceuticals (18 Desbrosses

Street, New York, New York 10013, USA). M M is a scientific consultant for

Bayer Pharma AG (Germany) and Eli Lilly (Indianapolis, Indiana, USA).

Funding

This work was supported by PRIN (Programmi di ricerca di Rilevante

Interesse Nazionale, protocol no. 20099BXMJH 002) and FIRB (Fondo per gli

investimenti alla ricerca di base, protocol no. 2010RBFR10VJ56 002), both

funds from the Italian Minister of University, Research and Instruction, by

Under40-Young Investigators funds from the Italian Minister of Health

(grant no. GR2008-1137632), and by a scientific grant from Intercept

Pharmaceuticals (18 Desbrosses Street, New York, New York 10013, USA).

References

Adorini L, Pruzanski M & Shapiro D 2012 Farnesoid X receptor targeting to

treat nonalcoholic steatohepatitis. Drug Discovery Today 17 988–997

(Review). (doi:10.1016/j.drudis.2012.05.012)

Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA,

Fruchart JC, James WP, Loria CM, Smith SC Jr et al. 2009 Harmonizing

the metabolic syndrome: a joint interim statement of the International

Diabetes Federation Task Force on Epidemiology and Prevention;

National Heart, Lung, and Blood Institute; American Heart Association;

World Heart Federation; International Atherosclerosis Society; and

International Association for the Study of Obesity. Circulation 120

1640–1645. (doi:10.1161/CIRCULATIONAHA.109.192644)

Bostrom P, Andersson L, Rutberg M, Perman J, Lidberg U, Johansson BR,

Fernandez-Rodriguez J, Ericson J, Nilsson T, Boren J et al. 2007 SNARE

proteins mediate fusion between cytosolic lipid droplets and are

implicated in insulin sensitivity. Nature Cell Biology 9 1286–1293.

(doi:10.1038/ncb1648)

Brasaemle DL 2007 Thematic review series: adipocyte biology. The perilipin

family of structural lipid droplet proteins: stabilization of lipid droplets

and control of lipolysis. Journal of Lipid Research 48 2547–2559.

(doi:10.1194/jlr.R700014-JLR200)

Cariou B, van Harmelen K, Duran-Sandoval D, van Dijk TH, Grefhorst A,

Abdelkarim M, Caron S, Torpier G, Fruchart JC, Gonzalez FJ et al. 2006

The farnesoid X receptor modulates adiposity and peripheral insulin

sensitivity in mice. Journal of Biological Chemistry 281 11039–11049.

(doi:10.1074/jbc.M510258200)

Cipriani S, Mencarelli A, Palladino G & Fiorucci S 2010 FXR activation

reverses insulin resistance and lipid abnormalities and protects against

liver steatosis in Zucker (fa/fa) obese rats. Journal of Lipid Research 51

771–784. (doi:10.1194/jlr.M001602)

Cusi K 2012 Role of obesity and lipotoxicity in the development of

nonalcoholic steatohepatitis: pathophysiology and clinical impli-

cations. Gastroenterology 142 711–725.e6. (doi:10.1053/j.gastro.2012.

02.003)

De Lean A, Munson PJ & Rodbard D 1978 Simultaneous analysis of families

of sigmoidal curves: applications to biossay, radioligand assay, and

physiological dose–response curves. American Journal of Physiology 235

97–102.

Despres JP & Lemieux I 2006 Abdominal obesity and metabolic syndrome.

Nature 444 881–887. (doi:10.1038/nature05488)

Donnelly KL, Smith CI, Schwarzenberg SJ, Jessurun J, Boldt MD & Parks EJ

2005 Sources of fatty acids stored in liver and secreted via lipoproteins

in patients with nonalcoholic fatty liver disease. Journal of Clinical

Investigation 115 1343–1351. (doi:10.1172/JCI 200523621)

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

Farrell GC 2009 The liver and the waistline: fifty years of growth. Journal of

Gastroenterology and Hepatology 24 S105–S118. (doi:10.1111/j.1440-

1746.2009.06080.x)

Filippi S, Vignozzi L, Morelli A, Chavalmane AK, Sarchielli E, Fibbi B, Saad F,

Sandner P, Ruggiano P, Vannelli GB et al. 2009 Testosterone partially

ameliorates metabolic profile and erectile responsiveness to PDE5

inhibitors in an animal model of male metabolic syndrome. Journal of

Sexual Medicine 6 3274–3288. (doi:10.1111/j.1743-6109.2009.01467.x)

Fiorucci S, Cipriani S, Baldelli F & Mencarelli A 2010 Bile

acid-activated receptors in the treatment of dyslipidemia and related

disorders. Progress in Lipid Research 49 171–185. (doi:10.1016/j.plipres.

2009.11.001)

Foster LJ, Yaworsky K, Trimble WS & Klip A 1999 SNAP23 promotes insulin-

dependent glucose uptake in 3T3-L1 adipocytes: possible interaction

with cytoskeleton. American Journal of Physiology 276 C1108–C1114.

Franck N, Stenkula KG, Ost A, Lindstrom T, Stralfors P & Nystrom FH 2007

Insulin-induced GLUT4 translocation to the plasma membrane is

blunted in large compared with small primary fat cells isolated from the

same individual. Diabetologia 50 1716–1722. (doi:10.1007/s00125-007-

0713-1)

Fu M, Wang C, Li Z, Sakamaki T & Pestell RG 2004 Minireview: cyclin D1:

normal and abnormal functions. Endocrinology 145 5439–5447

(Review). (doi:10.1210/en.2004-0959)

Goossens GH 2007 The role of adipose tissue dysfunction in the

pathogenesis of obesity-related insulin resistance. Physiology & Behavior

94 206–218. (doi:10.1016/j.physbeh.2007.10.010)

Gustafson B, Gogg S, Hedjazifar S, Jenndahl L, Hammarstedt A & Smith U

2009 Inflammation and impaired adipogenesis in hypertrophic obesity

in man. American Journal of Physiology. Endocrinology and Metabolism

297 E999–E1003 (Review). (doi:10.1152/ajpendo.00377.2009)

Jacobsson B & Smith U 1972 Effect of cell size on lipolysis and antilipolytic

action of insulin in human fat cells. Journal of Lipid Research 13

651–656.

Kawanishi M, Tamori Y, Okazawa H, Araki S, Shinoda H & Kasuga M 2000

Role of SNAP23 in insulin-induced translocation of GLUT4 in 3T3-L1

adipocytes. Mediation of complex formation between syntaxin4 and

VAMP2. Journal of Biological Chemistry 275 8240–8247. (doi:10.1074/

jbc.275.11.8240)

Larter CZ, Chitturi S, Heydet D & Farrell GC 2010 A fresh look at NASH

pathogenesis. Part 1: the metabolic movers. Journal of Gastroenterology

and Hepatology 25 672–690. (doi:10.1111/j.1440-1746.2010.06253.x)

Laurencikiene J, Skurk T, Kulyte A, Heden P, Astrom G, Sjolin E, Ryden M,

Hauner H & Arner P 2011 Regulation of lipolysis in small and large fat

cells of the same subject. Journal of Clinical Endocrinology and Metabolism

12 E2045–E2049. (doi:10.1210/jc.2011-1702)

Ma K, Saha PK, Chan L & Moore DD 2006 Farnesoid X receptor is essential

for normal glucose homeostasis. Journal of Clinical Investigation 116

1102–1109. (doi:10.1172/JCI25604)

Maneschi E, Morelli A, Filippi S, Cellai I, Comeglio P, Mazzanti B, Mello T,

Calcagno A, Sarchielli E, Vignozzi L et al. 2012 Testosterone treatment

improves metabolic syndrome-induced adipose tissue derangements.

Journal of Endocrinology 215 347–362. (doi:10.1530/JOE-12-0333)

Morelli A, Comeglio P, Filippi S, Sarchielli E, Cellai I, Vignozzi L,

Yehiely-Cohen R, Maneschi E, Gacci M, Carini M et al. 2012

Testosterone and farnesoid X receptor agonist INT-747 counteract high

fat diet-induced bladder alterations in a rabbit model of metabolic

syndrome. Journal of Steroid Biochemistry and Molecular Biology 132

80–92. (doi:10.1016/j.jsbmb.2012.02.007)

Morelli A, Comeglio P, Filippi S, Sarchielli E, Vignozzi L, Maneschi E,

Cellai I, Gacci M, Lenzi A, Vannelli GB et al. 2013 Mechanism of action

of phosphodiesterase type 5 inhibition in metabolic syndrome-

associated prostate alterations: an experimental study in the rabbit.

Prostate 73 428–441. (doi:10.1002/pros.22584)

Mudaliar S, Henry RR, Sanyal AJ, Morrow L, Marschall HU, Kipnes M,

Adorini L, Sciacca CI, Clopton P, Castelloe E et al. 2013 Efficacy and

safety of the farnesoid X receptor agonist obeticholic acid in patients

Published by Bioscientifica Ltd.

JournalofEndocrinology

Research E MANESCHI, L VIGNOZZI and others FXR signaling improves adiposetissue function

218 :2 231

with type 2 diabetes and nonalcoholic fatty liver disease.

Gastroenterology In press. (doi:10.1053/j.gastro.2013.05.042)

Olefsky JM 1976 The effects of spontaneous obesity on insulin binding,

glucose transport, and glucose oxidation of isolated rat adipocytes.

Journal of Clinical Investigation 57 842–851. (doi:10.1172/JCI108360)

O’Rourke RW, White AE, Metcalf MD, Olivas AS, Mitra P, Larison WG,

Cheang EC, Varlamov O, Corless CL, Roberts CT Jr et al. 2011 Hypoxia-

induced inflammatory cytokine secretion in human adipose tissue

stromovascular cells. Diabetologia 54 1480–1490. (doi:10.1007/s00125-

011-2103-y)

Pellicciari R, Fiorucci S, Camaioni E, Clerici C, Costantino G, Maloney PR,

Morelli A, Parks DJ & Willson TM 2002 6a-Ethyl-chenodeoxycholic

acid (6-ECDCA), a potent and selective FXR agonist endowed with

anticholestatic activity. Journal of Medicinal Chemistry 45 3569–3572.

(doi:10.1021/jm025529g)

Pellicciari R, Gioiello A, Macchiarulo A, Thomas C, Rosatelli E, Natalini B,

Sardella R, Pruzanski M, Roda A, Pastorini E et al. 2009 Discovery of

6a-ethyl-23(S)-methylcholic acid (S-EMCA, INT-777) as a potent and

selective agonist for the TGR5 receptor, a novel target for diabesity.

Journal of Medicinal Chemistry 52 7958–7961. (doi:10.1021/jm901390p)

Renga B, Mencarelli A, Vavassori P, Brancaleone V & Fiorucci S 2010

The bile acid sensor FXR regulates insulin transcription and secretion.

Biochimica et Biophysica Acta 1802 363–372. (doi:10.1016/j.bbadis.

2010.01.002)

Rizzo G, Disante M, Mencarelli A, Renga B, Gioiello A, Pellicciari R &

Fiorucci S 2006 The farnesoid X receptor promotes adipocyte

differentiation and regulates adipose cell function in vivo. Molecular

Pharmacology 70 1164–1173. (doi:10.1124/mol.106.023820)

Rizzo G, Passeri D, De Franco F, Ciaccioli G, Donadio L, Rizzo G, Orlandi S,

Sadeghpour B, Wang XX, Jiang T et al. 2010 Functional characterization

of the semisynthetic bile acid derivative INT-767, a dual farnesoid X

receptor and TGR5 agonist. Molecular Pharmacology 78 617–630.

(doi:10.1124/mol.110.064501)

Salans LB & Dougherty JW 1971 The effect of insulin upon glucose

metabolism by adipose cells of different size. Influence of cell lipid

and protein content, age, and nutritional state. Journal of Clinical

Investigation 50 1399–1410. (doi:10.1172/JCI106623)

Salans LB, Knittle JL & Hirsch J 1968 The role of adipose cell size and

adipose tissue insulin sensitivity in the carbohydrate intolerance of

human obesity. Journal of Clinical Investigation 47 153–165.

(doi:10.1172/JCI105705)

Salans LB, Bray GA, Cushman SW, Danforth E Jr, Glennon JA, Horton ES &

Sims EA 1974 Glucose metabolism and the response to insulin by

human adipose tissue in spontaneous and experimental obesity.

Effects of dietary composition and adipose cell size. Journal of Clinical

Investigation 53 848–856. (doi:10.1172/JCI107625)

Samocha-Bonet D, Chisholm DJ, Tonks K, Campbell LV & Greenfield JR

2012 Insulin-sensitive obesity in humans – a ‘favorable fat’ phenotype?

Trends in Endocrinology and Metabolism 23 116–124. (doi:10.1016/j.tem.

2011.12.005)

Sarruf DA, Iankova I, Abella A, Assou S, Miard S & Fajas L 2005 Cyclin D3

promotes adipogenesis through activation of peroxisome proliferator-

activated receptor g. Molecular and Cellular Biology 25 9985–9995.

(doi:10.1128/MCB.25.22.9985-9995.2005)

Sato H, Genet C, Strehle A, Thomas C, Lobstein A, Wagner A, Mioskowski C,

Auwerx J & Saladin R 2007 Anti-hyperglycemic activity of a TGR5

http://joe.endocrinology-journals.org � 2013 Society for EndocrinologyDOI: 10.1530/JOE-13-0109 Printed in Great Britain

agonist isolated from Olea europaea. Biochemical and Biophysical Research

Communications 362 793–798. (doi:10.1016/j.bbrc.2007.06.130)

Sinal CJ, Tohkin M, Miyata M, Ward JM, Lambert G & Gonzalez F 2000

Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid

and lipid homeostasis. Cell 102 731–744. (doi:10.1016/S0092-

8674(00)00062-3)

Smith U 1971 Effect of cell size on lipid synthesis by human adipose tissue

in vitro. Journal of Lipid Research 12 65–70.

Smith CE & Ordovas JM 2012 Update on perilipin polymorphisms and

obesity. Nutrition Reviews 70 611–621 (Review). (doi:10.1111/j.1753-

4887.2012.00515.x)

Snel M, Jonker JT, Schoones J, Lamb H, de Roos A, Pijl H, Smit JW,

Meinders AE & Jazet IM 2012 Ectopic fat and insulin resistance:

pathophysiology and effect of diet and lifestyle interventions.

International Journal of Endocrinology 2012 983814. (doi:10.1155/2012/

983814)

Student AK, Hsu RY & Lane MD 1980 Induction of fatty acid synthetase

synthesis in differentiating 3T3-L1 preadipocytes. Journal of Biological

Chemistry 255 4745–4750.

Targher G & Byrne CD 2013 Clinical review: nonalcoholic fatty liver

disease: a novel cardiometabolic risk factor for type 2 diabetes and its

complications. Journal of Clinical Endocrinology and Metabolism 98

483–495. (doi:10.1210/jc.2012-3093)

Thomas C, Pellicciari R, Pruzanski M, Auwerx J & Schoonjans K 2008

Targeting bile-acid signalling for metabolic diseases. Nature Reviews.

Drug Discovery 7 678–693 (Review). (doi:10.1038/nrd2619)

Thomas C, Gioiello A, Noriega L, Strehle A, Oury J, Rizzo G, Macchiarulo A,

Yamamoto H, Mataki C, Pruzanski M et al. 2009 TGR5-mediated bile

acid sensing controls glucose homeostasis. Cell Metabolism 10 167–177.

(doi:10.1016/j.cmet.2009.08.001)

Timmons JA, Wennmalm K, Larsson O, Walden TB, Lassmann T, Petrovic N,

Hamilton DL, Gimeno RE, Wahlestedt C, Baar K et al. 2007 Myogenic

gene expression signature establishes that brown and white adipocytes

originate from distinct cell lineages. PNAS 104 4401–4406.

(doi:10.1073/pnas.0610615104)

Trayhurn P 2013 Hypoxia and adipose tissue function and dysfunction in

obesity. Physiological Reviews 93 1–21. (doi:10.1152/physrev.00017.

2012)

Vignozzi L, Morelli A, Filippi S, Comeglio P, Chavalmane AK, Marchetta M,

Toce M, Yehiely-Cohen R, Vannelli GB, Adorini L et al. 2011 Farnesoid

X receptor activation improves erectile function in animal models of

metabolic syndrome and diabetes. Journal of Sexual Medicine 8 57–77.

(doi:10.1111/j.1743-6109.2010.02073.x)

Vignozzi L, Morelli A, Sarchielli E, Comeglio P, Filippi S, Cellai I,

Maneschi E, Serni S, Gacci M, Carini M et al. 2012 Testosterone protects

from metabolic syndrome-associated prostate inflammation: an

experimental study in rabbit. Journal of Endocrinology 212 71–84.

(doi:10.1530/JOE-11-0289)

Virtue S & Vidal-Puig A 2010 Adipose tissue expandability, lipotoxicity

and the metabolic syndrome – an allostatic perspective. Biochimica

et Biophysica Acta 1801 338–349 (Review). (doi:10.1016/j.bbalip.2009.

12.006)

Zhang Y, Lee FY, Barrera G, Lee H, Vales C, Gonzalez FJ, Willson TM &

Edwards PA 2006 Activation of the nuclear receptor FXR improves

hyperglycemia and hyperlipidemia in diabetic mice. PNAS 103

1006–1011. (doi:10.1073/pnas.0506982103)

Received in final form 23 June 2013Accepted 7 June 2013Accepted Preprint published online 7 June 2013

Published by Bioscientifica Ltd.