Relationship between FGF21 and UCP1 levels under time-restricted feeding and high-fat diet

    Relationship between FGF21 and UCP1 levels under time-restricted feeding and high-fat diet Nava Chapnik, Yoni Genzer, Oren Froy PII: DOI: Reference:

S0955-2863(16)30651-9 doi: 10.1016/j.jnutbio.2016.10.017 JNB 7676

To appear in:

The Journal of Nutritional Biochemistry

Received date: Revised date: Accepted date:

25 April 2016 14 October 2016 18 October 2016

Please cite this article as: Chapnik Nava, Genzer Yoni, Froy Oren, Relationship between FGF21 and UCP1 levels under time-restricted feeding and high-fat diet, The Journal of Nutritional Biochemistry (2016), doi: 10.1016/j.jnutbio.2016.10.017

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Relationship between FGF21 and UCP1 levels under time-restricted feeding and high-fat diet

RI P

T

Nava Chapnik, Yoni Genzer, Oren Froy*

Institute of Biochemistry, Food Science and Nutrition, Robert H. Smith Faculty of Agriculture,

SC

Food and Environment, The Hebrew University of Jerusalem, Rehovot 76100, Israel

MA NU

Running Head: Relationship between FGF21 and UCP1

ED

Word count: 2,134

CE

PT

Keywords: FGF21; UCP1; adipose tissue; circadian clock; nutrition; metabolism

Oren Froy

AC

*Corresponding author:

Phone: 972-8-948-9746 Fax: 972-8-936-3208

E-mail: [email protected]

1

ACCEPTED MANUSCRIPT Abbreviation List BAT – Brown adipose tissue

T

FGF21 – Fibroblast growth factor 21

RI P

HFD – High-fat diet KD – Ketogenic diet

SC

LFD – Low-fat diet RF – Restricted feeding

MA NU

SCN – Suprachiasmatic nuclei UCP1 – Uncoupling protein 1

AC

CE

PT

ED

WAT – White adipose tissue

2

ACCEPTED MANUSCRIPT Abstract FGF21 (fibroblast growth factor 21) exhibits a circadian oscillation and its induction is critical

T

during fasting. When secreted by liver and skeletal muscle, FGF21 enhances thermogenic activity

RI P

in brown adipose tissue (BAT) by utilizing uncoupling protein 1 (UCP1) to dissipate energy as heat. Recently, it has been reported that UCP1 is not required for FGF21-mediated reduction in body

SC

weight or improvements in glucose homeostasis. As the relationship between FGF21 and UCP1 induction in tissues other than BAT is less clear, we tested the effect of restricted feeding (RF) and

MA NU

high dietary fat on FGF21 circadian expression and its correlation with UCP1 expression in liver and white adipose tissue (WAT). High dietary fat disrupted Fgf21 mRNA circadian oscillation, but increased its levels in WAT. RF led to increased liver FGF21 protein levels, whereas those of UCP1

ED

decreased. In contrast, WAT FGF21 protein levels increased under high-fat diet, whereas those of UCP1 decreased under RF. In summary, FGF21 exhibits circadian oscillation, which is disrupted

PT

with increased dietary fat. The relationship between FGF21 and UCP1 levels depends on the tissue

AC

CE

and the cellular energy status.

3

ACCEPTED MANUSCRIPT 1. Introduction Mammals have developed an endogenous circadian clock located in the suprachiasmatic nuclei

T

(SCN) of the anterior hypothalamus that respond to the environmental light-dark cycle. The SCN

RI P

receives light information from the retina and transmits synchronization cues via neuronal connections or circulating humoral factors to peripheral clocks, such as the liver, heart and lungs,

SC

regulating cellular and physiological functions [1, 2]. The clock mechanism in both SCN neurons and peripheral tissues includes CLOCK and BMAL1 (brain-muscle-Arnt-like 1) that heterodimerize

MA NU

and bind to E-box sequences to mediate transcription of a large number of genes, including Periods (Per1, Per2, Per3) and Cryptochromes (Cry1, Cry2) [3].

The circadian clock regulates metabolism by mediating the expression and/or activity of

ED

certain metabolic enzymes, hormones, and transport systems [4]. The rhythmic expression and activity of the metabolic pathways is mainly attributed to the robust and coordinated expression of

PT

clock genes in metabolic tissues, such as the liver, adipose tissue and muscle [5, 6]. Disruption of the coordination between the endogenous clock and the environment leads to a greatly attenuated

CE

diurnal feeding rhythm, hyperphagia and obesity, and the metabolic syndrome [7-10].

AC

FGF21 (fibroblast growth factor 21) belongs to a superfamily of FGF peptides without mitogenic effects [11, 12]. FGF21 exhibits a circadian oscillation with a peak of expression correlating with that of free fatty acids [13]. FGF21 induction is critical for lipolysis, gluconeogenesis, and ketogenesis during the prolonged adaptive response to fasting [14, 15]. FGF21 is also induced in response to cold stress [16, 17]. Under this condition, FGF21 acts in an autocrine/paracrine manner to engage thermogenic pathways in brown adipose tissue (BAT) and also to induce the browning of white adipose tissue (WAT) [17, 18]. When secreted by the liver and skeletal muscle, FGF21 enhances thermogenic activity in BAT in order to maintain body temperature in neonates [17]. To facilitate this, BAT utilizes uncoupling protein 1 (UCP1), located within the inner mitochondrial membrane, to dissipate energy as heat [19]. FGF21-mediated

4

ACCEPTED MANUSCRIPT increases in UCP1 levels have led to a widespread speculation that UCP1-dependent thermogenesis may drive the therapeutic actions of FGF21. Surprisingly, it has recently been reported that UCP1 is

T

not required for FGF21-mediated reduction in body weight or improvements in glucose

RI P

homeostasis [20, 21]. Although FGF21 actions appear to correlate with increased Ucp1 mRNA levels in brown adipose tissues and protein immunoreactivity in beige or brown fat [18, 22], the

SC

correlation between FGF21 and UCP1 induction in other tissues, such as white adipose tissue and liver is less clear. In this study, we tested the effect of restricted feeding (RF) and high dietary fat on

AC

CE

PT

ED

MA NU

FGF21 circadian expression and its relationship with UCP1 expression in liver and WAT.

5

ACCEPTED MANUSCRIPT 2. Materials and Methods 2.1. Animals, treatments, and tissues

T

Four-week-old male C57BL/6 mice were housed in a temperature- and humidity-controlled facility

RI P

(23–24°C, 60% humidity). Mice were entrained to a light-dark cycle of 12 h light and 12 h darkness (LD) for two weeks with food available ad libitum (AL). After two weeks, mice were fed AL low-

SC

fat diet (AL-LFD), AL high-fat diet (AL-HFD), restricted feeding (RF) LFD (RF-LFD), RF-HFD or ketogenic diet (KD, Harlan Laboratories Ltd., Jerusalem, Israel) for 8 weeks. Mice were placed

MA NU

together in cages and each group consisted of 24 mice. The HF diet was based on soybean oil and palm stearin (fatty acid composition: 0.3% C:12, 1.3% C:14, 55% C:16, 5.1% C:18, 29.5% C:18-1, 7.4% C:18-2, 0.7% C:18-3) and contained 22% w/w fat (42.1% of calories), 17.3% of calories from

ED

protein and 40.6% of calories from carbohydrates. KD was based on stearin (fatty acid composition: 0.1% C:12, 0.3% C:14, 14.9% C:16, 11.4% C:18, 25.44% C:18-1, 41.42% C:18-2, 4.7% C:18-3)

PT

and contained 67.4% w/w fat (90.5% of calories), 9.1% of calories from protein and 0.4% of calories from carbohydrates. The RF group was given food between ZT4 and ZT8 (ZT0 is the time

CE

of lights on). After 8 weeks, on the last day of the experiment, mice were fasted for 12 h and

AC

subsequently anesthetized and liver and white adipose tissue (WAT) were removed every 4 h around the circadian cycle in total darkness (DD) under dim red light to avoid the masking effect of light. Tissues were immediately frozen in liquid nitrogen and stored at -80°C until further analysis. Mice were humanely killed at the end of the experiment. The joint ethics committee (IACUC) of the Hebrew University and Hadassah Medical Center approved this study.

2.2. RNA extraction and quantitative real-time PCR RNA was extracted from tissues using TRI Reagent (Sigma). Total RNA was DNase I-treated using RQ1 DNase (Promega, Madison, WI, USA) and reverse-transcribed using qScript cDNA synthesis kit (Quanta BioSciences, Gaithersburg, MD, USA) and random hexamers (Promega). The reaction

6

ACCEPTED MANUSCRIPT was subjected to quantitative real-time PCR using primers spanning exon-exon boundaries and the ABI Prism 7300 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). Primers

T

were tested alongside the normalizing gene Actin. The fold change in target gene expression was

RI P

calculated by the 2-∆∆Ct relative quantification method (Applied Biosystems).

SC

2.3. Western blot analyses

Tissues were lysed in lysis buffer, as was described [23]. Samples were run on a 10% SDS

MA NU

polyacrylamide gel and transferred onto nitrocellulose membranes as was described [23]. Blots were incubated with FGF21 (Cat # AB171941, 1/1000 dilution) and UCP1 (Cat # AB10983, 1/1000 dilution) (Abcam, Cambridge, UK) antibodies and after several washes, with horseradish

ED

peroxidase-conjugated secondary antibody (Cat # 111-035-003, 1/1000 dilution, Jackson ImmunoResearch, West Grove, PA, USA). Anti-mouse antibody (Cat # 691001, 1/3000 dilution,

PT

MP Biomedicals, Solon, OH, USA) was used to detect actin, the loading control. The immune reaction was detected by enhanced chemiluminescence (Santa Cruz Biotechnologies, Santa Cruz,

AC

units.

CE

CA, USA). Finally, bands were quantified by scanning and densitometry and expressed as arbitrary

2.4. Statistical analyses

All results are expressed as means ±SE. Student's t-test with normal distribution was used for comparison between 2 group. Tukey’s honestly significant difference (HSD) was performed as a single-step multiple comparison procedure in conjunction with ANOVA for the evaluation of significant differences. Further analysis of circadian rhythmicity was performed using CircWave software (version 1.4) (Circadian Rhythm Laboratory, University of Groningen, Groningen, The Netherlands). CircWave is an F-tested forward harmonic regression procedure and automatically detects how many harmonics can be added by F-test criterion (step forward regression style). For all

7

ACCEPTED MANUSCRIPT analyses, the significance level was set at p<0.05. Statistical analysis was performed with JMP

AC

CE

PT

ED

MA NU

SC

RI P

T

(version 21) software (SAS Institute, Inc., Cary, NC, USA).

8

ACCEPTED MANUSCRIPT 3. Results We set out to measure the amplitude and levels of FGF21 and UCP1 protein under low-fat diet

T

(LFD), high-fat diet (HFD), ketogenic diet (KD) and time-restricted feeding (RF), a feeding

RI P

regimen with no caloric restriction, as we have previously shown [23]. As is shown in Figure 1, average food intake to total body mass ratio (g food/g BW) was 0.107± 0.002 and 0.111± 0.006 for

SC

the AL-LF and RF-LF group, respectively, yielding a ~103% food consumption of the RF group. Similarly, average food intake to total body mass ratio (g food/g BW) was 090.0± 0.000 and 0.0.0±

MA NU

0.000 for the AL-HF and RF-HF group, respectively, yielding a ..% food consumption of the RF group (Fig. 1). Except for KD, the HFD groups were not significantly different from the LFD

ED

groups in their food consumption (Fig. 1).

3.1. High dietary fat disrupts Fgf21 mRNA circadian oscillation

PT

Under AL-LFD, Fgf21 mRNA oscillated around the circadian cycle with a peak during the dark phase in both white adipose tissue (WAT) and liver (p<0.05, CircWave) (Fig. 2A,C). In the liver,

CE

AL-HFD (42% of calories from fat) blunted the amplitude and in WAT and liver advanced the

AC

expression of Fgf21 mRNA (Fig. 2A,C). Ketogenic diet (90% of calories from fat) led to disrupted cycling of Fgf21 mRNA in both WAT and liver (p>0.05, CircWave) (Fig. 2 A,C). RF advanced the expression of Fgf21 mRNA. Interestingly, the overall daily mRNA levels were lower under high dietary fat in the liver, but not in WAT (p<0.05, Tukey HSD) (Fig. 2B,D). Taken together, these results show that high dietary fat disrupts Fgf21 mRNA circadian oscillation, but increases its levels in WAT.

3.2. Time-RF induces FGF21 protein, but reduces UCP1 protein levels in liver Although at the protein levels FGF21 did not exhibit circadian oscillation under LFD, HFD led to increased amplitude (Fig. 3A). RF led to increased FGF21 protein levels compared with AL

9

ACCEPTED MANUSCRIPT regardless whether it was high-fat or low-fat diet (p˂0.05, Tukey HSD) (Fig. 3B). High-fat diet did not affect FGF21 protein levels, but KD slightly reduced FGF21 protein levels (p˂0.05, Tukey

T

HSD) (Fig. 3B). Interestingly, under RF, the levels of UCP1 were lower compared with the AL-

RI P

HFD and AL-LFD groups (p˂0.05, Tukey HSD) and the protein expression phase was delayed (Fig. 3C,D). Under KD, the expression phase of UCP1 was similar to that of AL-HFD (Fig. 3C).

SC

Taken together, these results show that liver FGF21 protein levels increase under RF, whereas those

MA NU

of UCP1 decrease.

3.3. Time-RF does not affect FGF21 protein, but reduces UCP1 protein levels in WAT FGF21 protein levels were higher under ad libitum high-fat diet (AL-HFD) compared to ad libitum

ED

low-fat diet (AL-LFD) (p˂0.05, Tukey HSD) (Fig. 4B). RF did not affect the levels of FGF21 protein (p>0.05, Tukey HSD). Under KD, FGF21 levels were similar to those under LFD (Fig. 4.

PT

B). Similarly to the liver, under RF, the levels of UCP1 were lower compared with the AL groups regardless whether it was high-fat or low-fat diet (p˂0.05, Tukey HSD) (Fig. 4D). KD led to

CE

increased UCP1 levels (Fig. 4D). Both FGF21 and UCP1 exhibited increased amplitude under RF-

AC

LF (Fig. 4A,C). Taken together, these results show that WAT FGF21 protein levels are increased under high fat diet, whereas those of UCP1 decrease under RF.

10

ACCEPTED MANUSCRIPT 4. Discussion 4.1. FGF21 circadian oscillation

T

Our results show that Fgf21 mRNA exhibits circadian oscillation under low-fat diet and this

RI P

oscillation is blunted with the increase of dietary fat. This effect was less pronounced at the protein level due most likely to differences between transcriptional regulation by the core clock mechanism

SC

and translational regulation. The oscillation of FGF21 at the protein and mRNA levels has been previously shown in both rodents and humans [13, 24, 25]. Our results also show that the more

MA NU

dietary fat in the diet the more blunted the circadian expression is in both liver and WAT. Indeed, it was recently reported that FGF21 receptor (Fgfr4) and Klb (β-klotho), important for FGF21 signaling, were downregulated in the liver, while Fgfr1, FGF21 receptor, was down-regulated in

ED

WAT of KD-fed mice [26]. The disrupted oscillation under high dietary fat is congruent with

PT

previous reports [27-29].

4.2. FGF1 levels and dietary fat

CE

The levels of FGF21 protein increase under fasting conditions and correlate with elevated free fatty

AC

acids [13, 30]. However, studies have shown that FGF21 levels also increased during obesity, which could indicate a state of FGF21 resistance [30]. However, our results show that high-fat diet led to reduced liver Fgf21 mRNA levels, as has been shown in some studies [31]. Other studies have shown increased FGF21 after long-term treatment with high-fat diet in liver, WAT and serum [32]. We could only detect increased Fgf21 mRNA and FGF21 protein levels in WAT as a result of high-fat diet. Our results are congruent with previous findings that report FGF21 increased expression in adipocytes in the fed state, a pattern opposite to that in hepatocytes [33].

4.3. FGF21 vs. UCP1 levels in liver and WAT

11

ACCEPTED MANUSCRIPT The correlation between increased FGF21 and UCP1 levels in BAT has been well documented in the literature [34]. WAT depots are able to convert to a "brown-like" state with prolonged cold

T

exposure or exposure to β-adrenergic compounds characterized by the appearance of pockets of

RI P

UCP1-positive adipocytes in a process involving FGF21 [18]. We show that FGF21 protein levels increased under HFD compared with LFD in WAT. In the liver neither FGF21 levels nor those of

SC

UCP1 were changed as a result of high-fat diet. Interestingly, RF, which allows 20 h of no feeding without caloric restriction showed decreased levels of UCP1 protein in both liver and WAT. These

MA NU

results indicate that time restricted feeding does not induce "browning" of WAT. In WAT, RF did not affect the levels of FGF21 protein and in the liver RF increased FGF21 levels. Altogether, these findings suggest a different regulation of FGF21 and UCP1 expression in the various tissues under

ED

different conditions. Indeed, it has been reported that the majority of FGF21-driven metabolic endpoints do not require UCP1, [21]. More specifically, it has been reported that the antiobesogenic

PT

effects of FGF21 are independent of brite adipocytes and that FGF21 is efficacious in Ucp1 null mice [20]. However, improvement in clearance of a glucose load by FGF21 is lost in the absence of

AC

CE

UCP1 [35].

4.4. FGF21 vs. UCP1 levels in liver and WAT under KD Under KD, the levels of FGF21 protein resembled those of the AL groups in the liver, and those of the LFD groups in WAT. Under KD, the levels of UCP1 protein resembled those of the RF groups in the liver, and exhibited high expression in WAT. This unique effect of KD may indicate the combination of the effect of the high dietary content as well as the low carbohydrate content, forcing a state of catabolism.

4.5. Conclusions

12

ACCEPTED MANUSCRIPT Our results show that FGF21 exhibits circadian oscillation, which is disrupted with increased dietary fat. The relationship between FGF21 and UCP1 levels depends on the tissue and the cellular

T

energy status. Further study is needed in order to delineate the role of FGF21 in different tissues and

RI P

under various metabolic situation.

SC

Acknowledgements

MA NU

This study was supported by the Israel Science Foundation (1044/12).

Conflict of Interest

AC

CE

PT

ED

All contributing authors report no conflict of interest.

13

ACCEPTED MANUSCRIPT References [1] Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, et al. Coordinated

T

transcription of key pathways in the mouse by the circadian clock. Cell. 2002;109:307-20.

RI P

[2] Reppert SM, Weaver DR. Coordination of circadian timing in mammals. Nature. 2002;418:93541.

SC

[3] Schibler U, Ripperger J, Brown SA. Peripheral circadian oscillators in mammals: time and food. J Biol Rhythms. 2003;18:250-60.

MA NU

[4] Froy O. Metabolism and circadian rhythms--implications for obesity. Endocr Rev. 2010;31:124.

[5] Lee C, Etchegaray JP, Cagampang FR, Loudon AS, Reppert SM. Posttranslational mechanisms

ED

regulate the mammalian circadian clock. Cell. 2001;107:855-67. [6] Zvonic S, Ptitsyn AA, Conrad SA, Scott LK, Floyd ZE, Kilroy G, et al. Characterization of

PT

peripheral circadian clocks in adipose tissues. Diabetes. 2006;55:962-70. [7] Sancar G, Brunner M. Circadian clocks and energy metabolism. Cellular and molecular life

CE

sciences : CMLS. 2014;71:2667-80.

AC

[8] Turek FW, Joshu C, Kohsaka A, Lin E, Ivanova G, McDearmon E, et al. Obesity and metabolic syndrome in circadian Clock mutant mice. Science. 2005;308:1043-5. [9] Tsai LL, Tsai YC, Hwang K, Huang YW, Tzeng JE. Repeated light-dark shifts speed up body weight gain in male F344 rats. Am J Physiol Endocrinol Metab. 2005;289:E212-7. [10] Rudic RD, McNamara P, Curtis AM, Boston RC, Panda S, Hogenesch JB, et al. BMAL1 and CLOCK, two essential components of the circadian clock, are involved in glucose homeostasis. PLoS Biol. 2004;2:e377. [11] Cuevas-Ramos D, Almeda-Valdes P, Aguilar-Salinas CA, Cuevas-Ramos G, Cuevas-Sosa AA, Gomez-Perez FJ. The role of fibroblast growth factor 21 (FGF21) on energy balance, glucose and lipid metabolism. Curr Diabetes Rev. 2009;5:216-20.

14

ACCEPTED MANUSCRIPT [12] Kliewer SA, Mangelsdorf DJ. Fibroblast growth factor 21: from pharmacology to physiology. Am J Clin Nutr. 2010;91:254S-7S.

T

[13] Yu H, Xia F, Lam KS, Wang Y, Bao Y, Zhang J, et al. Circadian rhythm of circulating

RI P

fibroblast growth factor 21 is related to diurnal changes in fatty acids in humans. Clin Chem. 2011;57:691-700.

SC

[14] Badman MK, Pissios P, Kennedy AR, Koukos G, Flier JS, Maratos-Flier E. Hepatic fibroblast growth factor 21 is regulated by PPARalpha and is a key mediator of hepatic lipid metabolism

MA NU

in ketotic states. Cell Metab. 2007;5:426-37.

[15] Inagaki T, Dutchak P, Zhao G, Ding X, Gautron L, Parameswara V, et al. Endocrine regulation of the fasting response by PPARalpha-mediated induction of fibroblast growth factor 21. Cell

ED

Metab. 2007;5:415-25.

[16] Chartoumpekis DV, Habeos IG, Ziros PG, Psyrogiannis AI, Kyriazopoulou VE, Papavassiliou

PT

AG. Brown adipose tissue responds to cold and adrenergic stimulation by induction of FGF21. Mol Med. 2011;17:736-40.

CE

[17] Hondares E, Iglesias R, Giralt A, Gonzalez FJ, Giralt M, Mampel T, et al. Thermogenic

AC

activation induces FGF21 expression and release in brown adipose tissue. J Biol Chem. 2011;286:12983-90.

[18] Fisher FM, Kleiner S, Douris N, Fox EC, Mepani RJ, Verdeguer F, et al. FGF21 regulates PGC-1alpha and browning of white adipose tissues in adaptive thermogenesis. Genes Dev. 2012;26:271-81. [19] Cannon B, Nedergaard J. Brown adipose tissue: function and physiological significance. Physiological reviews. 2004;84:277-359. [20] Veniant MM, Sivits G, Helmering J, Komorowski R, Lee J, Fan W, et al. Pharmacologic Effects of FGF21 Are Independent of the "Browning" of White Adipose Tissue. Cell Metab. 2015;21:731-8.

15

ACCEPTED MANUSCRIPT [21] Samms RJ, Smith DP, Cheng CC, Antonellis PP, Perfield JW, 2nd, Kharitonenkov A, et al. Discrete Aspects of FGF21 In Vivo Pharmacology Do Not Require UCP1. Cell reports.

T

2015;11:991-9.

RI P

[22] Coskun T, Bina HA, Schneider MA, Dunbar JD, Hu CC, Chen Y, et al. Fibroblast growth factor 21 corrects obesity in mice. Endocrinology. 2008;149:6018-27.

SC

[23] Sherman H, Frumin I, Gutman R, Chapnik N, Lorentz A, Meylan J, et al. Long-term restricted

J Cell Mol Med. 2011;15:2745-59.

MA NU

feeding alters circadian expression and reduces the level of inflammatory and disease markers.

[24] Foo JP, Aronis KN, Chamberland JP, Paruthi J, Moon HS, Mantzoros CS. Fibroblast growth factor 21 levels in young healthy females display day and night variations and are increased in

ED

response to short-term energy deprivation through a leptin-independent pathway. Diabetes Care. 2013;36:935-42.

PT

[25] Oishi K, Uchida D, Ishida N. Circadian expression of FGF21 is induced by PPARalpha activation in the mouse liver. FEBS Lett. 2008;582:3639-42.

CE

[26] Asrih M, Altirriba J, Rohner-Jeanrenaud F, Jornayvaz FR. Ketogenic Diet Impairs FGF21

AC

Signaling and Promotes Differential Inflammatory Responses in the Liver and White Adipose Tissue. PLoS One. 2015;10:e0126364. [27] Barnea M, Madar Z, Froy O. High-fat diet delays and fasting advances the circadian expression of adiponectin signaling components in mouse liver. Endocrinology. 2009;150:161-8. [28] Barnea M, Madar Z, Froy O. High-fat diet followed by fasting disrupts circadian expression of adiponectin signaling pathway in muscle and adipose tissue. Obesity (Silver Spring). 2010;18:230-8. [29] Kohsaka A, Laposky AD, Ramsey KM, Estrada C, Joshu C, Kobayashi Y, et al. High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metab. 2007;6:414-21.

16

ACCEPTED MANUSCRIPT [30] Mai K, Bobbert T, Groth C, Assmann A, Meinus S, Kraatz J, et al. Physiological modulation of circulating FGF21: relevance of free fatty acids and insulin. Am J Physiol Endocrinol Metab.

T

2010;299:E126-30.

RI P

[31] Hao L, Huang KH, Ito K, Sae-Tan S, Lambert JD, Ross AC. Fibroblast Growth Factor 21 (Fgf21) Gene Expression Is Elevated in the Liver of Mice Fed a High-Carbohydrate Liquid

SC

Diet and Attenuated by a Lipid Emulsion but Is Not Upregulated in the Liver of Mice Fed a High-Fat Obesogenic Diet. J Nutr. 2016;146:184-90.

MA NU

[32] Sun B, Yang G, Yang M, Liu H, Boden G, Li L. Long-term high-fat diet links the regulation of the insulin-sensitizing fibroblast growth factor-21 and visfatin. Cytokine. 2012;59:131-7. [33] Dutchak PA, Katafuchi T, Bookout AL, Choi JH, Yu RT, Mangelsdorf DJ, et al. Fibroblast

ED

growth factor-21 regulates PPARgamma activity and the antidiabetic actions of thiazolidinediones. Cell. 2012;148:556-67.

PT

[34] Straub L, Wolfrum C. FGF21, energy expenditure and weight loss - How much brown fat do you need? Mol Metab. 2015;4:605-9.

CE

[35] Kwon MM, O'Dwyer SM, Baker RK, Covey SD, Kieffer TJ. FGF21-Mediated Improvements

AC

in Glucose Clearance Require Uncoupling Protein 1. Cell reports. 2015;13:1521-7.

17

ACCEPTED MANUSCRIPT Figure legends Figure 1: Fgf21 mRNA circadian expression and average daily levels in liver and WAT of AL-

T

LFD-, AL-HFD- and KD-fed mice. A) Fgf21 mRNA circadian expression in the liver. B) Fgf21

RI P

mRNA daily average levels in the liver. C) Fgf21 mRNA circadian expression in WAT. D) Fgf21 mRNA daily average levels in WAT. Tissues were collected in total darkness every 4 h around the

SC

circadian cycle from mice fed low-fat diet, high-fat diet and ketogenic diet. Total RNA was extracted, reverse-transcribed and expression levels were determined by real-time PCR. Circadian

MA NU

expression is plotted so that 1 is the lowest point. Data are means ± SE; n = 4 for each time-point in each group. Different letters denote significant difference (p<0.05, Tukey’s HSD).

ED

Figure 2: Body weight and food consumption of mice under AL-LFD, RF-LFD, AL-HFD and RF-HFD. A) Mouse average body weight. B) Cumulative average daily food consumption per

PT

mouse. C) Average food consumption to body weight to ratio. D) Average food consumption to

CE

body weight0.75. n=24 mice in each group.

AC

Figure 3: FGF21 and UCP1 protein circadian expression and average daily levels in the liver of mice fed AL-LFD, RF-LFD, AL-HFD and RF-HFD. A) FGF21 protein circadian expression. B) FGF21 protein levels. C) UCP1 protein circadian expression. D) UCP1 protein levels. Tissues were collected in total darkness every 4 h around the circadian cycle. Protein levels were quantified by Western blotting. Circadian expression is plotted so that 1 is the lowest point. Data are means ± SE; n = 4. Different letters denote significant difference (p<0.05, Tukey’s HSD).

Figure 4: FGF21 and UCP1 protein circadian expression and average daily levels in WAT of mice fed AL-LFD, RF-LFD, AL-HFD and RF-HFD. A) FGF21 protein circadian expression. B) FGF21 protein levels. C) UCP1 protein circadian expression. D) UCP1 protein levels. Tissues were

18

ACCEPTED MANUSCRIPT collected in total darkness every 4 h around the circadian cycle. Protein levels were quantified by Western blotting. Circadian expression is plotted so that 1 is the lowest point. Data are means ± SE;

AC

CE

PT

ED

MA NU

SC

RI P

T

n = 4. Different letters denote significant difference (p<0.05, Tukey’s HSD).

19

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

Figure 1

20

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

Figure 2

21

AC

Figure 3

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

22

CE AC

Figure 4

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

23