ORIGINAL ARTICLE

Fibroblast Growth Factor-21 and the Beneficial Effectsof Long-Chain n-3 Polyunsaturated Fatty Acids

Joan Villarroya • Pavel Flachs • Ibon Redondo-Angulo •

Marta Giralt • Dasa Medrikova • Francesc Villarroya •

Jan Kopecky • Anna Planavila

Received: 23 October 2013 / Accepted: 27 August 2014 / Published online: 10 September 2014

� AOCS 2014

Abstract Long-chain n-3 polyunsaturated fatty acids (LC

n-3 PUFA) in the diet protect against insulin resistance and

obesity. Fibroblast growth factor-21 (Fgf21) is a hormonal

factor released mainly by the liver that has powerful anti-

diabetic effects. Here, we tested whether the beneficial

metabolic effects of LC n-3 PUFA involve the induction of

Fgf21. C57BL/6 J mice were exposed to an obesogenic,

corn-oil-based, high-fat diet (cHF), or a diet in which corn

oil was replaced with a fish-derived LC n-3 PUFA con-

centrate (cHF ? F) using two experimental settings: short-

term (3 weeks) and long-term treatment (8 weeks).

CHF ? F reduced body weight gain, insulinemia, and tri-

glyceridemia compared to cHF. cHF increased plasma

Fgf21 levels and hepatic Fgf21 gene expression compared

with controls, but these effects were less pronounced or

absent in cHF ? F-fed mice. In contrast, hepatic expres-

sion of peroxisome proliferator-activated receptor (PPAR)-

a target genes were more strongly induced by cHF ? F

than cHF, especially in the short-term treatment setting.

The expression of genes encoding Fgf21, its receptors, and

Fgf21 targets was unaltered by short-term LC n-3 PUFA

treatment, with the exception of Ucp1 (uncoupling protein

1) and adiponectin genes, which were specifically up-reg-

ulated in white fat. In the long-term treatment setting, the

expression of Fgf21 target genes and receptors was not

differentially affected by LC n-3 PUFA. Collectively, our

findings indicate that increased Fgf21 levels do not appear

to be a major mechanism through which LC n-3 PUFA

ameliorates high-fat-diet-associated metabolic disorders.

Keywords Fibroblast growth factor-21 � Long-chain n-3

polyunsaturated fatty acids

Abbreviations

Acox1 Acyl-coenzyme A oxidase

BAT Brown adipose tissue

CPT2 Carnitine palmitoyltransferase-2

DHA Docosahexaenoic acid

Ehhadh Bifunctional enzyme or enoyl-coenzyme A,

hydratase/3-hydroxyacyl coenzyme A

dehydrogenase

EPA Eicosapentaenoic acid

FGF21 Fibroblast growth factor-21

FGFR Fibroblast growth factor receptor

Slc2a1 Glucose transporter-1

Acadm Medium chain acyl-CoA dehydrogenase

Ppargc1a PPARgamma-coactivator-1aPPAR Peroxisome proliferator-activated receptor

PUFA Polyunsaturated fatty acids

Electronic supplementary material The online version of thisarticle (doi:10.1007/s11745-014-3948-x) contains supplementarymaterial, which is available to authorized users.

J. Villarroya � I. Redondo-Angulo � M. Giralt � F. Villarroya �A. Planavila (&)

Departament de Bioquimica i Biologia Molecular, Institut de

Biomedicina de la Universitat de Barcelona (IBUB), University

of Barcelona, Barcelona, Spain

e-mail: aplanavila@ub.edu

J. Villarroya

Hospital de la Santa Creu i Sant Pau, Barcelona, Spain

P. Flachs � D. Medrikova � J. Kopecky

Department of Adipose Tissue Biology, Institute of Physiology

of Academy of Sciences of the Czech Republic, v.v.i, Prague,

Czech Republic

I. Redondo-Angulo � M. Giralt � F. Villarroya � A. Planavila

CIBER Fisiopatologia de la Obesidad y Nutricion, Barcelona,

Spain

123

Lipids (2014) 49:1081–1089

DOI 10.1007/s11745-014-3948-x

UCP1 Uncoupling protein-1

WAT White adipose tissue

Introduction

Long-chain polyunsaturated fatty acids of the n-3 series

(LC n-3 PUFA) from fish are known to exert beneficial

effects on human health by improving lipid metabolism,

exerting anti-inflammatory effects, and acting to prevent

diabetes and cardiovascular disease. Studies have demon-

strated decreased adiposity of obese humans and improved

glucose metabolism in healthy lean individuals after die-

tary LC n-3 PUFA supplementation [1, 2]. Studies in

rodents fed a high-fat (HF) diet or a lipogenic sucrose-rich

diet have indicated that LC n-3 PUFA counteract the

development of obesity and decrease associated metabolic

disorders, such as diabetes and dyslipidemia [1]. The

beneficial effects of o LC n-3 PUFA include effects on

adipose tissues, such as the enhancement of mitochondrial

activity in white adipose tissue (WAT) [3, 4], and the

prevention of insulin resistance in the liver [5].

Fibroblast growth factor-21 (Fgf21) is an FGF family

member that acts as a hormone. It is mainly released by the

liver, although it may also be expressed and released by

other tissues such as WAT and brown adipose tissue

(BAT). Fgf21 has strong anti-diabetic and anti-obesity

effects [6]. In genetic and diet-induced rodent models of

obesity and type II diabetes, it has been shown to lower

glucose, increase insulin sensitivity, and it even moderates

the development of an obese condition [7, 8]. Most of these

effects of Fgf21 occur via induction of glucose uptake and

oxidation in adipose tissues [7], although promotion of

BAT-related thermogenic activity [9] and autocrine effects

in the liver have also been reported [10]. Fgf21 acts on

tissues mainly through interaction with FGFR1 or FGFR4,

two FGF receptors, in critical dependence with the co-

receptor b-Klotho, to mediate the effects of Fgf21 [11]. In

the liver, the Fgf21 gene is under the transcriptional control

of fatty acid derivatives that act via peroxisome prolifera-

tor-activated receptor-a (PPARa) [10, 12]. Fatty acids

originating either from WAT lipolysis, from HF diet in

adult mice, or milk lipids in suckling pups, have been

shown to induce hepatic Fgf21 gene expression and Fgf21

release in a PPARa-dependent manner [13].

Here, in an effort to determine whether Fgf21 could

form a mechanistic link in the beneficial metabolic effects

of LC n-3 PUFA, we examined the effects of a high-fat diet

supplemented with or without LC n-3 PUFA on Fgf21

levels, gene expression for Fgf21 and components of the

tissue Fgf21 response machinery, as well as overall sys-

temic changes.

Methods

Two experimental settings were established for determin-

ing the effects of LC n-3 PUFA-enriched high fat diets. In

the first (short-term, high-dose LC n-3 PUFA), 2-month-

old male (C57BL/6 J) mice were fed three types of diet

profiles: [1] control (C), a high-carbohydrate, low-fat chow

diet (lipid content 3 %; A04 diet, Panlab, Barcelona, Spain)

for 5 weeks; [2] cHF, a corn-oil–based high-fat diet (lipid

content 35 %) for 5 weeks; and [3] cHF ? F, an initial

2-week cHF diet followed by 3 further weeks of the cHF

diet supplemented with an LC n-3 PUFA concentrate

(46 % DHA, 14 % EPA; EPAX 1050 TG; EPAX, Aales-

und, Norway) that replaced 44 % of dietary lipids (spe-

cifically corn oil). This treatment strategy has been

previously described [14]. In the second set of experiments

(long-term, lower-dose LC n-3 PUFA concentrate), the

same three treatment group design was followed, but

treatments were maintained for 8 weeks; the two cHF

groups maintained a 35 % (wt/wt) total lipid content, but in

the cHF ? F group, the LC n-3 PUFA concentrate replaced

15 % of dietary lipids, with the remaining lipids being

corn-oil–based fat. This second experimental setting has

also been previously described [15]. All experiments were

performed in accordance with ECC directive 86/609/EEC

and were approved by the Animal Experimentation Ethics

Committee, University of Barcelona.

Mice were maintained at 22 �C on a 12-hour light–dark

cycle (light from 8:00 a.m.) with free access to water. Food

consumption and body weight were recorded twice a week.

Mice were sacrificed by decapitation between 8:00 and

10:00 a.m. under post-prandial (fed) conditions. Blood

glucose and triglyceride levels were measured using an

Accutrend device (Roche Diagnostics, Basel, Switzerland).

Leptin, insulin, interleukin-6 (IL-6), plasminogen activator

inhibitor type-1 (PAI-1), and resistin were quantified using

a multiplex system (Linco Research/Millipore, Saint

Charles, MO, USA) and Luminex100ISv2 equipment.

Enzyme-linked immunosorbent assays (ELISAs) were used

to quantify Fgf21 (Biovendor, Modrice, Czech Republic)

and adiponectin (Life Technologies/Invitrogen, Carlsbad,

CA, USA) protein levels.

RNA was extracted using the NucleoSpin RNAII kit

(Macherey–Nagel, Duren, Germany). Quantitative real-

time reverse transcription-polymerase chain reaction (RT-

PCR) analyses of mRNA expression were performed using

standardized TaqMan probes (see Supplemental Fig. 1;

Applied Biosystems, Foster City, CA, USA) following the

supplier’s methodology. The relative amount of mRNA in

each sample was normalized to that of the reference control

18S rRNA using the comparative (2�D CT ) method. The

data were analyzed by one-way analysis of variance

1082 Lipids (2014) 49:1081–1089

123

(ANOVA), followed by Bonferroni or Dunnet’s post hoc

tests, as appropriate (GraphPad Software Inc., San Diego,

CA, USA).

Results

With 5-week treatment, body weight was increased in cHF

mice relative to controls in association with an increase in

epididymal WAT mass (Table 1). In contrast, there was no

significant increase in body weight or epididymal WAT in

cHF ? F mice compared to controls. cHF-fed mice con-

sumed significantly more calories than mice fed the control,

high-carbohydrate diet, whereas caloric intake was similar

Table 1 Effects of short-term cHF and cHF ? F diet treatment on

body weight, food intake, tissue weights, and circulating levels of

metabolites, hormones, and cytokines

Control diet cHF diet cHF ? F

diet

Body weight (increase

due to diet, g)

0.7 ± 0.03 4.3 ± 0.9* 0.15 ± 0.04#

Food intake (kJ/day) 60 ± 3 85 ± 4* 83 ± 4*

Liver (g) 1.51 ± 0.11 1.23 ± 0.94 1.54 ± 0.76

Interscapular BAT (mg) 109 ± 9 113 ± 14 104 ± 6#

Epididymal WAT (mg) 455 ± 89 958 ± 207* 469 ± 54#

Blood levels of

Glucose (mg/dl) 125 ± 5 130 ± 2 133 ± 7

Triglycerides (mg/dl) 142 ± 9 203 ± 11* 73 ± 3*, #

Plasma levels of

Insulin (ng/ml) 1.07 ± 0.28 1.03 ± 0.62 0.29 ± 0.06*

Leptin (ng/ml 67 ± 0.36 2.97 ± 1.64 0.94 ± 0.59

PAI-I (ng/ml) 2.89 ± 1.71 0.61 ± 0.19 1.11 ± 0.28

Resistin (ng/ml) 0.41 ± 0.16 0.26 ± 0.06 0.34 ± 0.23

IL-6 (pg/ml) 7.57 ± 0.33 7.86 ± 1.09 4.31 ± 2.0

Adiponectin (lg/ml) 41.4 ± 7.2 57.4 ± 13 60.2 ± 5.4*

cHF 5 weeks of corn-oil–based high-fat diet; cHF ? F 2 weeks of

cHF diet followed by 3 further weeks of treatment with cHF diet

supplemented with LC n-3 PUFA (see ‘‘Methods’’). Data are

mean ± SEMs (n = 5–6 mice/group). Data were analyzed by one-

way ANOVA

* p \ 0.05 versus the control diet; # p \ 0.05, cHF ? F versus cHF

cFig. 1 Effects of short-term dietary treatment with LC n-3 PUFA on

Fgf21 protein levels in plasma and mRNA levels in the liver. Mice

were treated with a high-carbohydrate control diet (C) or a corn-oil–

based high-fat diet (cHF) for 5 weeks, or a corn-oil–based high-fat for

2 weeks with 3 further weeks of cHF diet supplemented with an LC

n-3 PUFA concentrate (cHF ? F). a Plasma Fgf21 levels. b mRNA

expression levels of Fgf21, Fgf21 receptors, and metabolic genes in

the liver. The results are expressed as mean ± SEMs (n = 5–6 mice/

group; *p \ 0.05 versus the control diet; #p \ 0.05, cHF ? F versus

cHF)

Lipids (2014) 49:1081–1089 1083

123

between mice in cHF and cHF ? F groups (Table 1). In this

short-term experimental setting, the cHF diet did not sig-

nificantly affect the levels of blood glucose, insulin, leptin,

PAI-1, resistin or IL-6, whereas it triggered a slight increase

in blood triglycerides. The cHF ? F diet, in contrast, was

associated with significant reductions in blood triglycerides

Fig. 2 Effects of short-term dietary treatment with LC n-3 PUFA on

mRNA expression levels in WAT and BAT. Mice were treated with a

high-carbohydrate control diet (C) or a corn-oil–based high-fat diet

(cHF) for 5 weeks, or a corn-oil–based high-fat for 2 weeks with 3

further weeks of cHF diet supplemented with an LC n-3 PUFA

concentrate (cHF ? F). mRNA expression levels of Fgf21, Fgf21

receptors, adipogenic, and thermogenic genes in epididymal WAT

(a) and interscapular BAT (b). mRNA levels are expressed as

mean ± SEMs (n = 5–6 mice/group; *p \ 0.05 versus the control

diet; #p \ 0.05, cHF ? F versus cHF)

1084 Lipids (2014) 49:1081–1089

123

and insulin levels, and a significant increase in adiponectin

levels relative to controls (Table 1), indicating that an LC

n-3 PUFA-enriched high-fat diet enhanced systemic insulin

sensitivity. Collectively, these results are consistent with

previous findings for these diets in mice [14].

After 5-week treatment, the plasma levels of Fgf21 were

significantly higher in cHF-treated mice than in controls.

Interestingly, cHF ? F treatment did not significantly

increase the plasma levels of Fgf21, which were interme-

diate between those of control and cHF-treated mice

(Fig. 1a). The cHF diet dramatically induced Fgf21 gene

expression in the liver relative to controls (Fig. 1b).

Although Fgf21 mRNA levels trended slightly higher in

cHF ? F-treated mice compared with controls, this dif-

ference did not reach significance. Notably, the levels of

hepatic Fgf21 mRNA in cHF ? F-treated mice were sig-

nificantly lower than those in cHF-treated mice. Since

PPARa plays a major role in controlling Fgf21 gene

expression in the liver [10, 12], we determined the

expression levels of known target genes of the PPARapathway and of PPARa (Ppara) itself. In contrast to our

findings for Fgf21 gene expression levels (Fig. 1b), the

hepatic mRNA expression for the liver bifunctional

enzyme, Ehhadh (enoyl-CoA, hydratase/3-hydroxyacyl

CoA dehydrogenase), a highly sensitive PPARa-target in

the liver [16], was moderately induced by cHF treatment

and very strongly induced by cHF ? F treatment. The

same tendency was observed for other PPARa-target

genes, including Acox1 (acyl-coenzyme A oxidase), Acadm

(medium chain acyl-CoA dehydrogenase) and Cpt2 (car-

nitine palmitoyltransferase-2), as well as for Ppara itself.

The hepatic expression levels of Fgfr4 and Klb (b-Klotho)

did not significantly differ according to diet.

We also analyzed the expression of Fgf21 and potential

target genes of Fgf21 in WAT and BAT tissues sensitive to

Fgf21 action [6, 13]. Overall, neither cHF nor cHF ? F

significantly altered the expression of Fgf21, Ppargc1a

(PPARc coactivator 1 alpha, also known as PGC-1a),

Slc2a1 (glucose transporter-1) or Pparg (PPARc) in WAT

(Fig. 2a) or BAT (Fig. 2b). Moreover, neither expression

of the components of the Fgf21 response machinery, Fgfr1

and Klb, nor expression of Fgf21 itself was altered in WAT

or BAT by the different dietary regimens. The expression

of Ucp1 and Adipoq (adiponectin) were differentially

altered in BAT and WAT by the diets. cHF caused a sig-

nificant induction of Ucp1 gene expression in BAT,

whereas cHF ? F led to intermediate levels of expression

(Fig. 2b). In WAT, only cHF ? F caused a significant

induction of Ucp1 and Adipoq gene expression (Fig. 2a).

When the dietary treatment with cHF was longer, it

caused a larger increase in body weight and adipose

accretion (Table 2), parallel induction in leptin levels and

marked hyperinsulinemia, as expected. In this second

experimental setting, body weight and WAT increases

were only moderately attenuated by supplementation of the

cHF ? F diet with LC n-3 PUFA. Hypertriglyceridemia

was ameliorated and plasma adiponectin concentrations

were significantly induced in cHF ? F-treated mice

Table 2 Effects of 8-week cHF and cHF ? F diets on body weight, food intake, tissue weights, and circulating levels of metabolites, hormones,

and cytokines

Control diet cHF diet cHF ? F diet

Body weight(increase due to diet, g) 5.7 ± 0.5 11.7 ± 1.3* 8.6 ± 1.0

Food intake (kJ/day) 62 ± 2 74 ± 2 70 ± 3

Liver (g) 1.68 ± 0.58 1.96 ± 0.99* 1.65 ± 10.9

Interscapular BAT (mg) 116.9 ± 11 291 ± 30* 227 ± 30*

Epididymal WAT (mg) 592 ± 75 2,504 ± 116* 2,134 ± 101*, #

Blood levels of

Glucose (mg/dl) 151 ± 6 146 ± 4 141 ± 3

Triglycerides (mg/dl) 133 ± 10 188 ± 21* 128 ± 13

Plasma levels of

Insulin (ng/ml) 1.61 ± 0.03 13.7 ± 2.92* 10.9 ± 2.15*

Leptin (ng/ml) 6.60 ± 1.81 56.7 ± 5.08* 64.5 ± 8.77*

PAI-I (ng/ml) 0.61 ± 0.08 1.78 ± 0.03* 1.59 ± 0.17*

Resistin (ng/ml) 2.47 ± 0.18 3.93 ± 0.36* 4.02 ± 0.30*

IL-6 (pg/ml) 6.3 ± 1.3 7.7 ± 1.3 9.4 ± 3.1

AdipoQ (lg/ml) 52.4 ± 1.7 57.5 ± 3.1 75.6 ± 5.7*

cHF corn-oil–based high-fat diet treatment (8 weeks); cHF ? F cHF diet supplemented with LC n-3 PUFA (8 weeks) (see ‘‘Methods’’). Data are

mean ± SEMs (n = 5–6 mice/group). Data were analyzed by one-way ANOVA. Data are mean ± SEMs (n = 5–6 mice/group)

* p \ 0.05 versus the control diet; # p \ 0.05, cHF ? F versus cHF

Lipids (2014) 49:1081–1089 1085

123

compared with controls. In this long-term, high-fat expo-

sure setting, the plasma levels of Fgf21 were significantly

higher in cHF-treated mice and much lower in cHF ? F-

treated mice (Fig. 3a). The expression of hepatic Fgf21

mRNA was strongly induced by the cHF diet and only very

moderately by cHF ? F diet; in fact, the hepatic Fgf21

mRNA levels in the cHF ? F-treated mice did not signif-

icantly differ from those of the high-carbohydrate diet-fed

controls (Fig. 3b). Effects of diet on PPARa and its target

genes in the long-term setting were less pronounced than

those in the short-term, high-dose LC n-3 PUFA diet,

shown above; the only significant increase found was that

for Ehhadh mRNA in the cHF ? F group. Again, there

were no significant effects on the expression of Fgfr4 or

Klb, encoding mediators of Fgf21 signaling in the liver. On

the other hand, changes in Fgf21 expression in adipose

tissues were similar in this second experimental setting;

Fgf21 mRNA levels in cHF-treated mice were significantly

higher in WAT (Fig. 4a) and BAT (Fig. 4b) relative to

controls, and were intermediate between cHF and control

levels in cHF ? F-treated mice. In WAT (Fig. 4a), the

expression of some putative FGF21 target genes such as

Ucp1 and Slc2a1 were unaltered, whereas Ppargc1a and

Pparg were reduced to a similar extent in cHF and

cHF ? F. In WAT, a similar decrease in Klb expression

was found in response to cHF and cHF ? F, whereas Fgfr1

expression was unaltered. In BAT (Fig. 4b), both cHF and

cHF ? F caused a similar induction of Ucp1 expression,

and a similar decrease in Klb expression.

Discussion

Long-standing evidence indicates that LC n-3 PUFA has

beneficial effects on metabolism, particularly with respect

to obesity-associated diseases. Multiple processes mediate

these effects, including enhanced oxidative activity in

WAT [14]. Fgf21 promotes beneficial effects on glucose

homeostasis, and its expression and release into blood is

induced by fatty acids. Since Fgf21 enhances glucose

uptake and oxidation in adipose tissues, we speculated that

the induction of Fgf21 by LC n-3 PUFA mediates the

beneficial metabolic effects of dietary LC n-3 PUFA.

b Fig. 3 Effects of long-term dietary treatment with LC n-3 PUFA on

Fgf21 protein levels in plasma and mRNA levels in the liver. Mice

were treated for 8 weeks with a high-carbohydrate control diet (C), a

corn-oil–based high-fat diet (cHF), or a corn-oil–based high-fat diet

supplemented with an LC n-3 PUFA concentrate (cHF ? F).

a Plasma Fgf21 levels. b mRNA expression levels of Fgf21, Fgf21

receptors, and metabolic genes in the liver. The results are expressed

as mean ± SEMs (n = 5–6 mice/group; *p \ 0.05 versus the control

diet; #p \ 0.05, cHF ? F versus cHF)

1086 Lipids (2014) 49:1081–1089

123

However, our current results do not support this hypothesis.

Instead, we found that a standard high-fat diet induced

Fgf21 levels and hepatic Fgf21 mRNA expression after

both short-term and long-term exposure more strongly than

did an LC n-3 PUFA-enriched diet, which leads to the

healthy metabolic profile typically associated with Fgf21

effects. High levels of plasma Fgf21 and increased hepatic

Fgf21 gene expression have been previously reported in

Fig. 4 Effects of long-term dietary treatment with LC n-3 PUFA on

mRNA expression levels in WAT and BAT. Mice were treated for

8 weeks with a high-carbohydrate control diet (C), a corn-oil–based

high-fat diet (cHF), or a corn-oil–based high-fat diet supplemented

with LC n-3 PUFA concentrate (cHF ? F). mRNA expression levels

of Fgf21, Fgf21 receptors, adipogenic, and thermogenic genes in

epididymal WAT (a) and interscapular BAT (b). mRNA levels are

expressed as mean ± SEMs (n = 5–6 mice/group; *p \ 0.05 versus

the control diet; #p \ 0.05, cHF ? F versus cHF)

Lipids (2014) 49:1081–1089 1087

123

long-term, conventional high-fat-diet mouse models of

obesity and metabolic syndrome [6, 17], as well as in obese

humans [18]. Our current demonstration that a conven-

tional HF diet preferentially increased Fgf21 levels and

induced hepatic Fgf21 expression relative to a HF diet

supplemented with LC n-3 PUFA, even after short-term

treatment (i.e., before the development of obesity), sug-

gests a relatively direct effect of dietary fatty acid com-

position on Fgf21 synthesis.

Our examination of changes in gene expression in WAT

and BAT, where Fgf21 is considered to play an autocrine

and perhaps endocrine role [6, 9], revealed marked dif-

ferences according to treatment duration. A short-term,

obesogenic, high-fat diet did not modify either Fgf21 gene

expression or the overall expression profile of Fgf21 target

genes; only some degree of Ucp1 induction was observed

in BAT in this setting. However, short-term treatment with

a diet containing high-dose LC n-3 PUFA specifically

induced the expression of Adipoq and Ucp1 in WAT; the

latter is considered a sign of ‘‘browning’’ of adipose tissue,

and both changes are considered to be metabolically ben-

eficial events. In contrast, long-term treatment with the

obesogenic, high-fat diet maximally induced Fgf21 gene

expression in WAT and BAT. In this experimental setting,

Fgf21 target genes were either similarly induced by both

types of high-fat diets (Ucp1 expression in BAT) or simi-

larly repressed (Ppargc1a, Ppparg and Adipoq in WAT).

Expression of the key component of Fgf21 responsiveness,

b-Klotho, was reduced by an obesogenic diet in WAT and

BAT, a previously reported event [17, 19]. LC n-3 PUFA

caused an identical downregulation. Collectively, these

data are consistent with previous reports indicating

impaired responsiveness to Fgf21 in adipose tissues in

response to an obesogenic, high-fat diet [19], and suggest

that LC n-3 PUFA does not prevent or reverse this effect.

This is partially in contrast with the short-time treatment

setting in which, despite no direct evidence, it cannot be

ruled out that high expression of Ucp1 and Adipoq, known

targets of Fgf21 in WAT [13, 20, 21], elicited by LC n-3

PUFA may reflect enhanced Fgf21 action, despite moder-

ate circulating levels of Fgf21. Perhaps the ability of short-

term, high-dose LC n-3 PUFA to totally preclude adipose

accretion may contribute to this distinct scenario relative to

that observed with long-term dietary treatments. On the

other hand, recent reports indicate that Fgf21 directly

promotes adiponectin secretion in adipose tissue [20, 21].

However, a divergence between Fgf21 levels and adipo-

nectin levels in response to a conventional high-fat diet

versus an LC n-3 PUFA-enriched diet was observed. It is

therefore unlikely that the increase in adiponectin levels

induced by an LC n-3 PUFA-enriched diet is mediated by

Fgf21. A similar inference applies to the induction of

Adipoq gene expression in adipose tissue, noted above.

On the other hand, data from humans and rodent models

have indicated a strong association between FGF21 levels

and hepatic stress conditions, ranging from hepatic stea-

tosis to hepatocarcinogenesis [22–24]. Thus, the lack of

Fgf21 induction in LC n-3 PUFA-treated mice could reflect

the hepato-protective role of this diet relative to the con-

ventional high-fat diet. In this sense, reduced hepatic Fgf21

expression in response to LC n-3 PUFA would be consis-

tent with the protective effect of LC n-3 PUFA against the

development of hepatic steatosis in response to a high-fat

diet [5].

Another conclusion that might be drawn from our

study is that the diet-induced change in hepatic Fgf21

gene expression is at least partially independent of the

PPARa-mediated pathway. The target genes of PPARawere more intensely induced in the livers of cHF ? F-fed

mice than in mice on a standard high-fat diet, especially

in the setting of short-term, high-dose LC n-3 PUFA

treatment. This is consistent with previous reports

obtained in vitro, indicating that polyunsaturated fatty

acids preferentially activate PPARa compared with

unsaturated or mono-unsaturated fatty acids that are

present in conventional high-fat diets [25]. Hepatic Fgf21

gene expression exhibited the opposite behavior, indicat-

ing that signals mediated by a high-fat diet unrelated to

PPARa may also be involved in the induction of Fgf21 in

the liver of high-fat-diet-fed mice.

In summary, present findings indicate that increased

Fgf21 levels and actions do not appear to be a major

mechanism through which LC n-3 PUFA ameliorates high-

fat-diet-associated metabolic disorders, and mechanisms

other than PPARa-mediated signaling are likely to mediate

Fgf21 regulation in response to fat components in diet.

Acknowledgments Supported by Ministerio de Ciencia e Innova-

cion (SAF2011-23636) and Instituto de Salud Carlos III (PI11/00376)

Spain; and the Czech Science Foundation (13-00871S), Czech

Republic. IR-A was supported by a pre-doctoral fellowship from

Gobierno Vasco (Programa de Formacion de Investigadores del

DEUI).

Conflict of interest The authors declare that they have no com-

peting interests.

References

1. Flachs P, Rossmeisl M, Bryhn M, Kopecky J (2009) Cellular and

molecular effects of n-3 polyunsaturated fatty acids on adipose

tissue biology and metabolism. Clin Sci (Lond) 116:1–16

2. Mori TA, Bao DQ, Burke V, Puddey IB, Watts GF, Beilin LJ

(1999) Dietary fish as a major component of a weight-loss diet:

effect on serum lipids, glucose, and insulin metabolism in over-

weight hypertensive subjects. Am J Clin Nutr 70:817–825

3. Flachs P, Horakova O, Brauner P, Rossmeisl M, Pecina P,

Franssen-van HN et al (2005) Polyunsaturated fatty acids of

1088 Lipids (2014) 49:1081–1089

123

marine origin upregulate mitochondrial biogenesis and induce

beta-oxidation in white fat. Diabetologia 48:2365–2375

4. Nisoli E, Tonello C, Cardile A, Cozzi V, Bracale R, Tedesco L

et al (2005) Calorie restriction promotes mitochondrial biogenesis

by inducing the expression of eNOS. Science 310:314–317

5. Jelenik T, Rossmeisl M, Kuda O, Jilkova ZM, Medrikova D, Kus

V et al (2010) AMP-activated protein kinase alpha2 subunit is

required for the preservation of hepatic insulin sensitivity by n-3

polyunsaturated fatty acids. Diabetes 59:2737–2746

6. Dutchak PA, Katafuchi T, Bookout AL, Choi JH, Yu RT, Man-

gelsdorf DJ et al (2012) Fibroblast growth factor-21 regulates

PPARgamma activity and the antidiabetic actions of thiazolid-

inediones. Cell 148:556–567

7. Kharitonenkov A, Shiyanova TL, Koester A, Ford AM, Mica-

novic R, Galbreath EJ et al (2005) FGF-21 as a novel metabolic

regulator. J Clin Invest 115:1627–1635

8. Coskun T, Bina HA, Schneider MA, Dunbar JD, Hu CC, Chen Y

et al (2008) Fibroblast growth factor 21 corrects obesity in mice.

Endocrinology 149:6018–6027

9. Hondares E, Iglesias R, Giralt A, Gonzalez FJ, Giralt M, Mampel

T et al (2011) Thermogenic activation induces FGF21 expression

and release in brown adipose tissue. J Biol Chem 286:12983–

12990

10. Inagaki T, Dutchak P, Zhao G, Ding X, Gautron L, Parameswara

V et al (2007) Endocrine regulation of the fasting response by

PPARalpha-mediated induction of fibroblast growth factor 21.

Cell Metab 5:415–425

11. Ogawa Y, Kurosu H, Yamamoto M, Nandi A, Rosenblatt KP,

Goetz R et al (2007) BetaKlotho is required for metabolic activity

of fibroblast growth factor 21. Proc Natl Acad Sci USA

104:7432–7437

12. Badman MK, Pissios P, Kennedy AR, Koukos G, Flier JS,

Maratos-Flier E (2007) Hepatic fibroblast growth factor 21 is

regulated by PPARalpha and is a key mediator of hepatic lipid

metabolism in ketotic states. Cell Metab 5:426–437

13. Hondares E, Rosell M, Gonzalez FJ, Giralt M, Iglesias R, Vil-

larroya F (2010) Hepatic FGF21 expression is induced at birth via

PPARalpha in response to milk intake and contributes to ther-

mogenic activation of neonatal brown fat. Cell Metab

11:206–212

14. Flachs P, Ruhl R, Hensler M, Janovska P, Zouhar P, Kus V et al

(2011) Synergistic induction of lipid catabolism and anti-

inflammatory lipids in white fat of dietary obese mice in response

to calorie restriction and n-3 fatty acids. Diabetologia

54:2626–2638

15. Horakova O, Medrikova D, van Schothorst EM, Bunschoten A,

Flachs P, Kus V et al (2012) Preservation of metabolic flexibility

in skeletal muscle by a combined use of n-3 PUFA and rosig-

litazone in dietary obese mice. PLoS ONE 7:e43764

16. Purushotham A, Schug TT, Xu Q, Surapureddi S, Guo X, Li X

(2009) Hepatocyte-specific deletion of SIRT1 alters fatty acid

metabolism and results in hepatic steatosis and inflammation.

Cell Metab 9:327–338

17. Diaz-Delfin J, Hondares E, Iglesias R, Giralt M, Caelles C, Vil-

larroya F (2012) TNF-alpha represses beta-Klotho expression and

impairs FGF21 action in adipose cells: involvement of JNK1 in

the FGF21 pathway. Endocrinology 153:4238–4245

18. Zhang X, Yeung DC, Karpisek M, Stejskal D, Zhou ZG, Liu F

et al (2008) Serum FGF21 levels are increased in obesity and are

independently associated with the metabolic syndrome in

humans. Diabetes 57:1246–1253

19. Fisher FM, Chui PC, Antonellis PJ, Bina HA, Kharitonenkov A,

Flier JS et al (2010) Obesity is a fibroblast growth factor 21

(FGF21)-resistant state. Diabetes 59:2781–2789

20. Lin Z, Tian H, Lam KS, Lin S, Hoo RC, Konishi M et al (2013)

Adiponectin mediates the metabolic effects of FGF21 on glucose

homeostasis and insulin sensitivity in mice. Cell Metab

17:779–789

21. Li H, Fang Q, Gao F, Fan J, Zhou J, Wang X et al (2010)

Fibroblast growth factor 21 levels are increased in nonalcoholic

fatty liver disease patients and are correlated with hepatic tri-

glyceride. J Hepatol 53:934–940

22. Dushay J, Chui PC, Gopalakrishnan GS, Varela-Rey M, Crawley

M, Fisher FM et al (2010) Increased fibroblast growth factor 21 in

obesity and nonalcoholic fatty liver disease. Gastroenterology

139:456–463

23. Yang C, Lu W, Lin T, You P, Ye M, Huang Y et al (2013)

Activation of Liver FGF21 in hepatocarcinogenesis and during

hepatic stress. BMC Gastroenterol 13:67

24. Kliewer SA, Sundseth SS, Jones SA, Brown PJ, Wisely GB,

Koble CS et al (1997) Fatty acids and eicosanoids regulate gene

expression through direct interactions with peroxisome prolifer-

ator-activated receptors alpha and gamma. Proc Natl Acad Sci

USA 94:4318–4323

25. Galman C, Lundasen T, Kharitonenkov A, Bina HA, Eriksson M,

Hafstrom I et al (2008) The circulating metabolic regulator

FGF21 is induced by prolonged fasting and PPARalpha activa-

tion in man. Cell Metab 8:169–174

Lipids (2014) 49:1081–1089 1089

123