High-intensity interval versus moderate-intensity continuous training: Superior metabolic benefits in diet-induced obesity mice

Life Sciences 191 (2017) 122–131

Contents lists available at ScienceDirect

Life Sciences journal homepage: www.elsevier.com/locate/lifescie

High-intensity interval versus moderate-intensity continuous training: Superior metabolic benefits in diet-induced obesity mice

MARK

Ningning Wanga, Yang Liub, Yanan Mab, Deliang Wena,b,⁎ a b

School of Public Health, Dalian Medical University, Dalian, Liaoning, China School of Public Health, China Medical University, Shenyang, Liaoning, China

A R T I C L E I N F O

A B S T R A C T

Keywords: Exercise intervention Obesity Lipid metabolism Insulin resistance Brown adipose tissue

Aims: Exercise is beneficial in obesity, however, the debate about the value of high-intensity interval training (HIIT) vs. moderate-intensity continuous training (MICT) has been long lasting. Therefore, here we have compared the possible beneficial effects of two different exercise training regimes in a mouse model of diet-induced obesity (DIO). Materials and methods: Following 7 wk. on high fat diet (HFD), ten-week-old male ICR mice (n = 30) were assigned to HIIT, distance-matched MICT or remained sedentary for the next 8 constitutive weeks while maintaining the dietary treatments. Age-matched sedentary mice with standard diet were used as a control (n = 10). Exercise was performed on a motorized treadmill for 5 days a week. Key findings: Both modes of exercise ameliorated adiposity and related metabolic dysfunction induced by HFD and sedentary lifestyle, while mice following HIIT exhibited significantly lower body weight, percentage of fat mass and smaller adipocyte size. HIIT was more favorable in preventing liver lipid accumulation by restoring mRNA levels of genes involved in hepatic lipogenesis (SREBP1, ACC1, FAS) and β-oxidation (PPARα, CPT1a, HAD). In addition, HIIT was more efficient in mitigating adipose tissue inflammation and insulin insensitivity, partly dependent on abrogating phosphorylation of JNK/IRS1 (Ser307) pathway. Moreover, only HIIT led to pronounced beige adipocyte recruitment in inguinal subcutaneous adipose tissue. Significance: We conclude that HIIT contribute a more favorable regulation of metabolic dysfunctions in DIO mice compared with MICT.

1. Introduction

driver of exercise participation and adherence [11], as HIIT involves alternating short bursts of high intensity exercise with recovery periods or light exercise [12]. In contrast, some other evidences supported similar health benefits of HIIT and traditional endurance exercise [13,14], even argued that HIIT may not be safe and tolerable [15]. Therefore, elucidating which exercise regime improves the metabolic status of individuals with obesity is essential. Exercise training improves whole-body energy metabolism largely attributed to adaptations in skeletal muscle [16–18]; however, training also affects many other tissues, including liver and adipose tissue. Exercise has yielded improvements in dyslipidemia of individuals affected by obesity, primarily through restoring the gene expression of molecules related to fat oxidation and lipogenesis [19,20]. Hitherto the effect of HIIT on hepatic lipid metabolism at transcriptional and protein level remains largely unexamined. In addition, adipose tissue is one of the most important organs linked to obesity-associated IR and chronic inflammation, since low-grade inflammation caused by an

Obesity, which is associated with cardiometabolic dysfunction, diabetes, respiratory disease, certain types of cancer, and osteoarthritis, has become one of the most serious threats to human health [1–3]. Lifestyle interventions against the obesogenic environment including unhealthy diet and sedentary behavior is deemed to be the most important strategy in tackling obesity [3]. Traditionally, the moderate-intensity continuous training (MICT) has been the most common type of exercise recommended to improve body composition, cardiorespiratory fitness, insulin resistance (IR), and lipid profile [4,5]. However, many people fail to accomplish the “traditional training programs”, ascribing to “lack of time” or “lack of enjoyment”. Accumulating studies have suggested high-intensity interval training (HIIT) yield more favorable results in weight loss, metabolic and cardiovascular status improvement than those with MICT [6–10]. Moreover, HIIT is perceived to be more enjoyable, serving as a stronger

⁎ Corresponding author at: Dalian Medical University, 9 W Lvshun South Road, Dalian, Liaoning Province, China; China Medical University, No.77 Puhe Road, Shenyang North New Area, Shenyang, Liaoning Province, China. E-mail address: [email protected] (D. Wen).

http://dx.doi.org/10.1016/j.lfs.2017.08.023 Received 15 May 2017; Received in revised form 19 August 2017; Accepted 22 August 2017 Available online 24 August 2017 0024-3205/ © 2017 Elsevier Inc. All rights reserved.

Life Sciences 191 (2017) 122–131

N. Wang et al.

2.4. Glucose tolerance tests (GTTs) and insulin tolerance tests (ITTs)

accumulation of adipose tissue and ectopic fat contributes to the pathogenesis of IR [21]. Although emerging evidences has proven that HIIT can improve insulin sensitivity [7,13,22], controversy lies in whether improvements are attributable to the exercise training per se or fat loss. Thus, it remains to be better understood the molecular responses to HIIT that incorporates inflammatory signaling and insulin sensitivity pathways in the adipose tissue. Despite the competence of exercise on expending energy stored in white adipose tissue (WAT) is such a cliche, recent studies focusing on the role of exercise in the activation of brown adipose tissue (BAT) shed lights on new strategies combating obesity [23–25], since activating BAT relates to stimulation of resting energy expenditure, nonshivering thermogenesis [26], and improvements in glucolipid homeostasis [27,28]. Also, the possibility of increasing energy expenditure by inducing a brown-like metabolic phenotype in WAT (a process nowadays called ‘browning’) is of great therapeutic interest. However, the kind of exercise prescription potentially recruiting brown and beige adipocytes remains undetermined. The present study aimed to clarify how HIIT would influence obesity-related physiological variables, and compare changes in these adaptations with traditional MICT in a mouse model of high fat diet (HFD) induced obesity. We proposed a mechanistic link between HIIT and advantageous metabolic homeostasis, providing theoretical basis for development of new and improved options to prevent and treat obesity and its related complications.

After 15 weeks of feeding and training, mice were fasted overnight (20:00–8:00) with free access to drinking water. A baseline blood sample (fasting blood glucose, FBG) was collected from the tail of fully conscious mice followed by an oral gavage of glucose (2.0 g/kg bw), and blood was taken from the tail at 15, 30, 60, 90 and 120 min postgavage. For ITTs, mice were fasted for 5 h (9:00–14:00) and insulin injection (0.75 unit/kg bw) (FosunPharm) was administered by intraperitoneal injection. Blood samples were collected as pre-described. The blood glucose concentration was determined with a glucometer (Onetouch Ultra, Johnson). GTTs and ITTs were manipulated 48 h after treadmill running to minimize the acute insulin sensitizing actions of exercise. To normalize for differences in basal glucose concentrations, these data were displayed as the area under the curves (AUCs) with subtraction of basal glucose concentrations. 2.5. Sample preparation At the end of the experiment, mice were sacrificed after 12 h of fasting. Liver, retroperitoneal WAT (rWAT), epididymal WAT (eWAT), inguinal subcutaneous white adipose tissue (scWAT), interscapular brown adipose tissue (iBAT), and gastrocnemius muscle were removed and weighed. Blood samples were collected and serum was separated by centrifugation (3000 rpm, 15 min). All samples were quick-frozen in liquid nitrogen and stored at − 80 °C for further use.

2. Material and methods

2.6. Biochemistry assay

2.1. Ethics statement

The serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), total cholesterol (TC), triglyceride (TG), high-density-lipoprotein cholesterol (HDL-c), and low-density lipoprotein cholesterol (LDL-c) were measured by auto chemistry analyzer (Hitachi7600–210, Japan),and high-sensitivity C-reactive protein (hsCRP) by automatic special protein analyzer (Siemens BNTII, German) according to the manufacturer's instructions. IL-6, adiponectin were analyzed using commercial kits (Merck Millipore) according to the manufacturer's standards. Fasting insulin (FINS) levels were analyzed using a mouse insulin ELISA kit (Thermo). Homeostasis model assessment of insulin resistance (HOMA-IR, an indicator of systemic insulin resistance) = FBG (mmol/L) × FINS (mU/L)/22.5. TG levels in liver were analyzed using a commercial kit (A110-1, Nanjing Jiancheng Bioengineering Institute, China). Briefly, small portions of liver tissue (50 mg) were collected and homogenized in ice-cold 100% ethanol (450 mL). After centrifugation, the supernatant was collected for analysis based on the glycerol lipase oxidase (GPO-PAP) method. Samples were reacted with the mixture from the kit and were incubated at 37 °C for 10 min, and absorbance at 510 nm was read with a microplate reader (Spectra MR, Dynex Technologies, USA).

This study was approved by the Animal Ethics Committee of the Institute of Genome Engineered Animal Models for Human Disease, Dalian Medical University (Permit Number: SYXK (liao) 2013–0006). All the experimental methods were carried out in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health. 2.2. Animals and diets For diet-induced obesity (DIO) phenotype, 3-weeks old male ICR mice (purchased from Institute of Genome Engineered Animal Models for Human Disease of Dalian Medical University, China) were fed with HFD (45%kcal fat, MD12032, Medicience Ltd.) for 7 weeks. Then the DIO mice were randomly assigned to different interventions (n = 10 in each group): HIIT, MICT or sedentary lifestyle (SED), and continued to consume the HFD. Meanwhile, lean control mice (CON) (n = 10) fed a standard control diet (10%kcal fat, MD12031, Medicience Ltd.) over the entire period were also included. Five mice per cage with access to food and purified water ad libitum were maintained on a 12/12 h light/ dark cycle with lights on at 08: 00 in a temperature (23 ± 2 °C) and humidity (60%) controlled room.

3. Histological analysis 3.1. Hematoxylin and eosin (H & E) staining

2.3. Exercise protocol Parts of liver and eWAT were excised, washed with ice-cold PBS, and fixed in 10% formalin. After being dehydrated in a grade alcohol series and embedded in paraffin wax, sections of tissue (thickness of 4–5 μm) were prepared and stained with H & E for histopathology and visualized by an Olympus BX63 microscope and pictured by Image Pro Plus7.0 software. Adipocyte size was calculated from three randomly selected fields of view for each animal as Singh et al. described [30], using the National Institutes of Health Image J software.

Exercise was performed on a motored mice treadmill (FT-200, Techman Soft) at 25° inclination 5 days/week (Monday to Friday) for 8 weeks, according to a protocol slightly modified from that described by Kemi et al. [29]. Both groups started with a warm-up at 5 m/min, where after HIIT consisted 10 bouts of 4 min high-intensity (85–90% VO2max) treadmill running, interspersed by 2 min active rest (5 m/min); whereas MICT consisted of distance-matched continuous running, corresponding to 65–70% of VO2max. The pace during HIIT and MICT was increased gradually from 16 to 26 m/min and 9 to 13 m/min over 8 weeks respectively (detailed in supplementary information, Table S1 and S2).

3.2. Immunohistochemistry (IHC) analysis Paraffin-embedded adipose tissues (iBAT, scWAT, and eWAT) were 123

Life Sciences 191 (2017) 122–131

N. Wang et al.

uncoupling protein-1 (UCP1, U6382, Sigma), PR domain containing 16 (PRDM16, ab106410, Abcam) or internal control β-actin (TA-09, ZSGBBIO), GAPDH (10494-1-AP, Proteintech). Blots were then incubated with relevant secondary antibodies, peroxidase-conjugated goat antimouse IgG (ZB-2305, ZSGB-BIO) or peroxidase-conjugated goat antirabbit IgG (ZB-2301, ZSGB-BIO) for 2 h. Bands were detected with the enhanced chemiluminescence detection systeme (KGP1127, Keygen Biotech) and Bio-Rad ChemiDoc™ MP imaging system. Relative abundance was measured with Image J software.

performed by using the DAB Detection Kit (SP-9000-D, ZSGB-BIO) according to the manufacturer's instructions. Sections were deparaffinized in xylene, hydrated in 100%, 95%, 85% and 75% ethanol, and rinsed in water before heat-mediated antigen retrieval in 10 mM pH 6.0 sodium citrate buffer. Quenching of endogenous peroxidases was performed using 3% H2O2 followed by incubations in nonspecific staining blockers for 15 min at 37 °C and overnight incubations with antibody against UCP1 (U6382, Sigma) at a dilution of 1:100 at 4 °C. Sections were then incubated in biotinylated secondary antibody at 37 °C for 30 min followed by avidin and biotinylated HRP (1:1) mixed solution incubations for 15 min at 37 °C. Afterward, the sections were visualized with DAB and counterstained with hematoxylin. Antigen distribution was examined under light microscope, and quantified (threshold area) by Image J software.

5. Quantitative reverse transcription PCR (qRT-PCR) Total RNA was extracted from frozen liver tissue with the use of RNAiso Plus (TaKaRa) according to the manufacturer's specifications. The integrity and quality of the purified RNA were analyzed by measurement of the A260/A280 ratio. RNA was reverse transcribed into cDNA using 5 × All-In-One RT MasterMix (ABM). For the qRT-PCR procedure, 100 ng of cDNA and 0.4 μM of each primer were used in 20 μL volume system containing SYBR Premix Ex Taq™ II (TaKaRa) and in a Rotor-Gene® Real-Time PCR instrument (Qiagen). All the samples were analyzed in triplicate. The gene-specific primer (ABM) sequences are listed in Table S3. The thermal cycling program was 95 °C for 30s, followed by 40 cycles of 95 °C for 5 s, 58 °C for 20 s. Following PCR, the melting curve was completed (95 °C for 1 s, 65 °C for 15 s) to ensure that only one PCR product was amplified per reaction. GAPDH was used as an endogenous control. The relative expression of the qRT-PCR products was determined by the ΔΔCt method.

3.3. Immunofluorescence (IF) analysis Sections of eWAT were also prepared for IF to assess macrophage infiltration. Tissue slides were deparaffinized in xylene, hydrated through graded ethanol series, and rinsed in PBS before heat-mediated antigen retrieval. Slides were permeabilized with 0.1% Triton X-100, quenched of endogenous peroxidases with 3% H2O2, and blocked with 5% BSA in PBS-T for 1 h, followed by incubation overnight with primary antibody for F4/80 (ab6640, Abcom) at a dilution of 1:200 in 1% BSA/PBS-T at 4 °C. Slides were then incubated with the Alex Fluor 488 Goat Anti-Mouse IgG (SA00006-1, Proteintech) labeled with FITC for 1 h. To visualize the nuclei, slides were fixed in 4′,6-diamidino-2-phenylindole (DAPI, Roche) at a dilution of 2.5 μg/mL for 20 min. Fluorescent antigen distribution was examined with a fluorescence microscope, and the intensity of fluorescence (threshold area) was quantified by Image J software.

6. Statistical analysis The data were expressed as the means ± standard error of mean (SEM) and analyzed using the SPSS 20.0 statistical software (SPSS Inc. USA). The comparisons between groups were analyzed using ANOVA followed by least-significant difference post hoc analysis, and P ≤ 0.05 was considered statistically significant.

3.4. Transmission electron microscopy (TEM) analysis Tissue samples were cut into small sections (1mm3) and fixed in 2.5% glutaraldehyde, post-fixed in 1% osmium tetroxide (OsO4), dehydrated through a graded ethanol series. After treatment of propylene oxide, the sections were embedded in epon. Ultra-thin (60 nm) sections of each sample were mounted on copper grids, then stained with lead citrate and 2% uranyl acetate routinely. For the morphometric analysis of liver sections, total mitochondria in three micrographs per group were counted, mitochondrial size and number were analyzed by Image J software; in view of the limited mitochondrial content in adipose tissue, mitochondrial number was counted in all fields of each section. Electron micrographs were taken with JEM 1400 transmission electron microscope at 80 kV.

7. Results 7.1. HIIT mice exhibited drastically reduced body weight, fat mass, and adipocyte size of WAT As illustrated in Fig. 1A & B, initial body weight (BW) of mice fed with HFD was significantly higher than that of CON group but did not differ between HFD groups. At individual time points, the BW of HIIT mice was lower than the SED mice and the difference was found to be statistically significant through 8 wk. of experiment. Weight gain of the training mice was significantly less than sedentary counterparts, in spite of higher total caloric intake (Table 1). In relative terms, HIIT induced weight loss more notable than MICT (p < 0.01). In addition, there was a slight weight rebound from week 6 to week 8 in two training groups, especially for the MICT group. The gross appearance of visceral WAT (vWAT, indicates the sum of rWAT and eWAT) showed a corresponding result (Fig. 1C): HIIT induced a striking reduction of fat mass compared with SED, while MICT exerted a smaller extent of reduction. Fig. 1 D & E shows fat mass and adiposity index of HIIT mice are 45% and 53% of those found in the MICT mice (both p < 0.01). As expected, HFD induced an evident adipocyte hypertrophy when compared to CON mice, in which almost the whole cell was occupied by one large lipid droplet, while cytoplasm was essentially undetectable. On average, the size of the adipocytes of HIIT mice was smaller than that of the SED and MICT mice (Fig. 1F). Morphometric analysis (Fig. 1G) confirmed these findings. Together, these results reveal that both HIIT and MICT are efficient in preventing HFD-induced weight and fat increase, albeit more energy intake. In particular, HIIT has a more profound impact than MICT.

4. Western blotting Total proteins from adipose tissues were extracted using the Minute™ total protein extraction kit (AT-022, Invent Biotechnologies), and proteins from liver/skeletal muscle were extracted with RIPA lysis buffer (20–188, Millipore). The concentration of protein extracts was quantified using Pierce™ BCA Protein Assay Kit (Thermo). Equal amounts of protein from each sample were separated by SDS-PAGE, and electrotransferred onto a polyvinylidene fluoride membrane. After being blocked with 10% non-fat milk for 1 h at 37 °C, the blots were incubated with appropriate diluted primary antibodies against sterol regulatory element-binding proteins-1 (SREBP1, ab28481, Abcam), peroxisome proliferator-activated receptor-alpha (PPARα, ab8934, Abcam), JNK1 (ab110724, Abcam), phospho-SAPK/JNK (Thr183/ Tyr185) (#4668, CST), IKKα/β (wl01900, WanleiBio), insulin receptor substrate-1 (IRS1, #2382, CST), phospho-IRS1 (Ser307) (#2381, CST), AKT1 (10176-2-AP, Proteintech), phospho-Akt (Ser473) (#4060, CST), glucose transporter-4 (GLUT4, wl02425, WanleiBio) (ab188317, Abcam), TNFα (wl01581, WanleiBio), IL-1β (wl00891, WanleiBio), 124

Life Sciences 191 (2017) 122–131

N. Wang et al.

Fig. 1. HIIT reduced body weight (BW), fat mass, and adipocyte size of white adipose tissue (WAT). (A) BW evolution and (B) BW change of mice that were sedentary or trained during 8 weeks. n = 10 per group. (C) Representative pictures of mice sedentary or trained, showing enormous discrepancy in body size (Left) and visceral WAT pads (Right) as indicated by arrows. (D and E) Quantification of total fat pads weight and adiposity index. n = 10 per group. (F) Representative hematoxylin and eosin (H & E) staining of cross-sections of epididymal WAT (eWAT) isolated from mice sedentary or trained. The adipocyte size was significantly larger in mice fed with HFD than CON group, whereas it appeared markedly shrunken in HIIT group compared with that of SED mice, but differences were not notable in MICT group. (G) Quantification of average lipid droplet size of eWAT from mice of all groups. n = 3 per group. Images are shown at 40 × magnification. Scale bars: 200 μm. Data are presented as mean ± SEM; *p ≤ 0.05, **p ≤ 0.01 vs CON; †p ≤ 0.05, ††p ≤ 0.01 vs SED; #p ≤ 0.05, ##p ≤ 0.01 HIIT vs MICT by ANOVA; CON, control; HFD, high fat diet; SED, sedentary behavior; HIIT, high intensity interval training; MICT, moderate intensity continuous training.

contained distorted cristae hardly been identified. In contrast, higher density (p < 0.01 vs. SED and vs. MICT) and size (p < 0.05 vs. SED and vs. MICT) of mitochondria were observed in liver cell of HIIT mice, with the morphology of mitochondria restored into elliptical or circular bursa, and amount of lipid droplets and ultrastructural damage to the cells were markedly decreased. Hepatic fat accumulation seems to be triggered by changes in gene expression of molecules related to lipid metabolism. RT-PCR analysis showed a mild decrease of gene expression of lipogenic transcription factor SREBP1, as well as of its downstream targets, ACC1, FAS and SCD1, in SED compared with CON mice (Fig. 2F). These HFD-induced altered gene expressions were totally prevented by HIIT with the exception of SCD1, meanwhile by MICT with the exception of SREBP1. Expressions of oxidative transcription PPARα and its downstream targets, CPT1a and HAD levels also tended to decline in SED compared with CON group. These HFD-induced altered gene expressions related to fat oxidation were totally prevented by HIIT but not MICT. At the protein level (Fig. 2H–J), PPARα expression showed a similar profile to its transcriptional level; however, SREBP1 was significantly up-regulated in SED group and was down-regulated effectively by training, and still the effect of HIIT was more remarkable in comparison with MICT (p < 0.05). Thus, HIIT as an alternative strategy, abrogates HFD-induced whole-body and hepatic lipodystrophy, consistent with

7.2. HIIT decreased serum and liver lipid accumulation, alleviated ultrastructural damage to hepatocytes caused by HFD, and altered expression of molecules related to hepatic lipid metabolism As illustrated in Table 1, serum ALT and TG content of two training groups were significantly lower than that observed in SED group. TC and LDL-c level of HIIT mice were significantly lower than that of SED mice, but did not differ between MICT and SED mice. Regarding the liver as a major organ where lipid uptake and utilization occur, H & E staining of liver sections (Fig. 2A) was performed: SED group exhibited a substantial amount of lipid deposition, with disorganized hepatic cord and cellular swelling, which were markedly alleviated in the HIIT and MICT groups. Moreover, HIIT almost restored the morphology to the same as normal controls. Biochemical analysis of liver TG content showed consistent results with morphological evaluation (Fig. 2B). Mitochondrion plays an important role in lipid metabolism, therefore we analyzed mitochondrion levels in liver tissues. As can be seen in Fig. 2C & E, the number of mitochondria tended to decline in SED group compared with CON (p > 0.05), with substantial amount of lipid droplets deposited, and characterized by other organelles damage such as dilated cisternae of rough endoplasmic reticulum. Importantly, most of the mitochondria appeared elongation into linear structure, and 125

Life Sciences 191 (2017) 122–131

N. Wang et al.

Table 1 Energy intake and serum parameters during or after 8 wk. of interventions. CON

Food intake (g/day/mouse) Energy intake (kcal/day/mouse) Water intake (mL/day/mouse) ALT (U/L) AST (U/L) TC (mmol/L) TG (mmol/L) HDL-c (mmol/L) LDL-c (mmol/L) hsCRP (ng/L) IL-6 (pg/mL) APN (μg/mL) FINS (mIU/L) FBG (mmol/L)

4.69 ± 0.21 18.05 ± 0.80 5.26 ± 0.30 85.33 ± 13.87 140.17 ± 17.20 5.23 ± 0.30 1.38 ± 0.06 2.94 ± 0.18 1.42 ± 0.09 0.59 ± 0.04 30.77 ± 1.28 1.15 ± 0.10 19.00 ± 0.93 6.25 ± 0.30

HFD SED

MICT

HIIT

4.52 ± 0.20 21.37 ± 0.94* 6.87 ± 0.31** 108.17 ± 15.77 126.00 ± 12.56 5.51 ± 0.38 2.05 ± 0.19** 2.57 ± 0.17 1.36 ± 0.13 1.26 ± 0.11** 61.75 ± 1.50** 1.14 ± 0.20 31.17 ± 0.84** 9.59 ± 0.37**

5.30 ± 0.18†† 25.08 ± 0.86†† 6.42 ± 0.19 51.00 ± 8.25†† 141.83 ± 19.66 4.70 ± 0.14 1.23 ± 0.10†† 2.78 ± 0.19 1.09 ± 0.05 0.97 ± 0.11 37.96 ± 1.19†† 1.08 ± 0.08 17.02 ± 0.67†† 7.48 ± 0.42††

5.11 ± 0.35† 24.17 ± 1.67† 6.86 ± 0.27 40.17 ± 4.58†† 106.67 ± 10.11 3.78 ± 0.34††# 1.35 ± 0.10†† 2.83 ± 0.22 0.87 ± 0.11†† 0.54 ± 0.06††## 28.58 ± 1.55††## 1.55 ± 0.14# 12.20 ± 0.92††## 5.46 ± 0.77††##

The results are presented as means ± SEM. Statistical analyses were conducted using one-way ANOVA followed by least significant post hoc tests. For all analyses, values of p ≤ 0.05 were considered statistically significant, n = 10 for energy intake and n = 6 for serum parameters analysis per group. *p ≤ 0.05, **p ≤ 0.01 vs CON group; †p ≤ 0.05, ††p ≤ 0.01 vs SED group; #p ≤ 0.05, ##p ≤ 0.01 HIIT vs MICT group; ALT, alanine aminotransferase; AST, aspartate aminotransferase; TC, total cholesterol; TG, triglyceride; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; hsCRP, high-sensitivity C-reactive protein; IL-6, interleukin-6; APN, adiponectin; FINS, fasting insulin; FBG, fating blood glucose.

Chronic inflammatory response of adipose tissue in obesity yielded potential mechanism underlying systemic IR, we therefore investigated the inflammatory responses at different levels (systemic and adipose tissue). As shown in Table 1, pro-inflammatory IL-6 and hsCRP level were significantly lower in HIIT group compared with SED and MICT groups. The content of anti-inflammatory adiponectin was comparable between sedentary and training mice, but it differed between HIIT and MICT mice (p < 0.05). To further investigate tissue specific inflammation, we measured the expression of macrophage activation marker (F4/80) and pro-inflammatory cytokines (TNFα, IL-1β) in the adipose tissue. In paralleled with serological findings, F4/80+ cells were immunostaining overtly in the SED group compared with CON, and down-regulated in two training group, especially in the HIIT group (Fig. 3F & G). In line with the grade of macrophage infiltration, we also observed an training-related decline in TNFα and IL-1β protein expressions compared to sedentary counterparts (p < 0.01) (Fig. 3H, N, O). To investigate a potential crosstalk in attenuation of inflammation and IR by exercise training, subsequently, we examined JNK and IKKβ (Fig. 3H), which might serve as an inhibitor of insulin signaling, by inducing inhibitory serine 307 phosphorylation of IRS1 [31]. As expected, JNK phosphorylation decreased in both HIIT and MICT group, when compared with SED mice (both p < 0.01), but was comparable between two training groups (Fig. 3L). No difference was observed in IKKβ protein level between diet matched sedentary and training groups (Fig. 3M).

mitochondrial function and lipid metabolic pathway adaptions. 7.3. HIIT ameliorated whole-body glucose homeostasis and inflammation, in accordance with modification of inflammatory and insulin signaling in adipose tissues To assess effects of training on glucose homeostasis, FINS, FBG and HOMA-IR were evaluated. The sedentary mice developed glycometabolic disorder under HFD stress, including a significant increase of FBG by 53% and FINS by 65% (Table 1), as well as HOMA-IR level by 1.5fold compared with CON mice (Fig. 3E). These indicators were significantly lower in two training groups, nevertheless, HIIT played a more profound impact than MICT (p < 0.01). Subsequently, we measured insulin sensitivity by performing ITTs and GTTs (Fig. 3A & C). Corresponding AUCs of ITTs indicated improvements in whole-body insulin sensitivity of two training groups compared with SED group (Fig. 3B), importantly, HIIT still exhibited a stronger effect than MICT (p < 0.05). However, surprisingly, glucose tolerance was comparable among all groups (Fig. 3D). Given the dramatic improvement in whole-body insulin sensitivity with training, we hypothesized that exercise training caused adaptations to the insulin signaling in insulin-responsive tissues. As shown in Fig. 3H–K, HFD induced a significant increase in serine phosphorylation of IRS1 (Ser307), and reduction in serine phosphorylation of Akt1 (Ser473) in vWAT, a more distal event in the insulin receptor signaling pathway, which both in turn inhibit insulin signaling. Surprisingly, GLUT4 which are responsible for insulin-stimulated glucose translocation to cell membrane, was increased in respond to HFD. Both HIIT and MICT demonstrated significant suppression of p-IRS (Ser307), activation of p-Akt (Ser473) and GLUT4 compared with SED group, and the value was significantly different between HIIT and MICT mice. Since skeletal muscle is the primary site of insulin (and exercise) -induced clearance of glucose in comparison to adipose tissue, we also measured markers of insulin signaling in skeletal muscle, and similar results were obtained (Fig. S1). HIIT was more advantageous in recovering pSer307-IRS and p-Ser473-Akt, whereas GLUT4 expression was more sensitive to MICT, indicating HIIT was more pronounced in ameliorating insulin sensitivity and MICT was more efficient in glucose disposal.

7.4. HIIT was particularly prone to enhancing thermogenic activity of BAT and inducing “browning” of subcutaneous adipose tissue Fig. 4A shows a striking hypertrophy and lighter color of iBAT in the sedentary HFD-fed mice. This was accompanied by a large number of unilocular adipocytes accumulation in iBAT with microscopic analysis (Fig. 4B). However, up-regulation of UCP1 + adipocytes has been shown to respond to HFD stress compared with CON mice fed with standard diet. Moreover, exercise training resulted in the emergence of more abundant small, multilocular UCP1 + adipocytes (Fig. 4B). The quantification analysis (Fig. 4C) was 2.0-fold and 1.7-fold higher in HIIT and MICT mice than the SED animals, respectively. Of note, HIIT was a stronger inducer of UCP1 expression in iBAT compared with 126

Life Sciences 191 (2017) 122–131

N. Wang et al.

Fig. 2. HIIT decreased liver lipid accumulation, alleviated ultrastructural damage to hepatocytes caused by HFD, and altered expression of molecules related to hepatic lipid metabolism. (A) Representative H & E staining of cross-sections of liver isolated from mice sedentary and trained. Extensive vacuolation was detected in hepatocytes of SED mice, which indicates lipid droplets accumulation (arrows), whereas it was reduced dramatically in the training groups, particularly in the HIIT group. Images are shown at 40 × magnification. Scale bars: 200 μm. (B) Quantification of liver triglyceride content. (C) Transmission electron micrographs of liver section from mice sedentary and trained. Hepatocytes of mice from SED contain abundance of lipid droplets and slightly declined mitochondria density, most of which appeared elongation into linear structure with distorted cristae hardly been identified. Other organelles damage such as numerous, dilated cisternae of endoplasmic reticulum (arrowheads) were also observed. Higher density of mitochondria emerged in two training group, especially in the hepatocytes of HIIT mice, with the morphology of mitochondria restored into elliptical or circular bursa, and amount of lipid droplets and ultrastructural damage to the cells was markedly decreased. Selected regions within the squares in low magnification images (15000 ×) are shown below in high magnification (30000 ×). L, lipid droplet; N, nucleus; Mit, mitochondrion. Scale bars: 0.5 μm. (D and E) Quantification of mitochondria size and number (expressed as the number of mitochondria per nucleus). (F and G) Normalized expression of lipogenesis- and β-oxidative-associated genes in liver tissue. (H) Protein expressions of SREBP1, PPARα, and internal control GAPDH in liver. (I and J) Quantification of proteins described in (H) with normalization to protein levels of GAPDH. Data are presented as mean ± SEM, n = 3 per group; *p ≤ 0.05, **p ≤ 0.01 vs CON; †p ≤ 0.05, ††p ≤ 0.01 vs SED; #p ≤ 0.05, ##p ≤ 0.01 HIIT vs MICT by ANOVA; CON, control; HFD, high fat diet; SED, sedentary behavior; HIIT, high intensity interval training; MICT, moderate intensity continuous training.

adipocytes and a marked increase in UCP1 content in this fat depot, while MICT mice showed no change in appearance relative to SED animals (Fig. 4F). The BAT-specific markers UCP1 and PRDM16 expression in scWAT was further confirmed at the protein level (Fig. 4G). Concomitant with morphology analysis, only HIIT significantly increased UCP1, as well as PRDM16 expression in scWAT (Fig. 4H & I). To

MICT (p < 0.05). The recruitment of brown-like adipocytes within WAT (known as ‘beige adipocytes’) is associated to increased energy expenditure and resistance to obesity [32]. Fig. 4D & E reveals an evident reduced volume of lipid droplets in inguinal scWAT of HIIT mice than SED and MICT mice. This was accompanied by presence of more multilocular

127

Life Sciences 191 (2017) 122–131

N. Wang et al.

Fig. 3. HIIT ameliorated whole-body glucose homeostasis and tissue specific inflammation, modified inflammatory and insulin signaling in adipose tissues. (A and C) Blood glucose concentration during insulin tolerance tests (ITTs) and oral glucose tolerance tests (GTTs). (B and D) The corresponding area under curves (AUCs) of ITTs and GTTs. n = 10 per group. (E) HOMA-IR = FBG (mmol/L) × FINS (mU/L)/22.5. n = 6 per group. HOMA-IR, homeostasis model assessment of insulin resistance; FINS, fasting insulin; FBG, fasting blood glucose. (F) Immunofluorescence (IF) analysis of macrophage infiltration in epididymal white adipose tissue (eWAT). Sections were exposed to F4/80 antibody (macrophage activation marker), each labeled with the fluorescent dye FITC. Left: fluorescent microscopy for FITC stain (green); Middle: DAPI nuclear stain (blue); Right: DAPI merged with FITC. F4/80 was overtly displayed by immunostaining in the SED group compared with the CON group, and down-regulated in two training group, especially in the HIIT. Images are shown at 20× magnification. Scale bar, 100 μm. n = 3 per group. DAPI, 4,6-diamidino2-phenylindole; FITC, fluorescein isothiocyanate. (G) Quantitative analysis of IF to determine F4/80 + adipocytes area in eWAT. (H) Protein expressions of inflammatory and insulin signaling pathway. For insulin-stimulated phosphorylated Akt (Ser473) and IRS1 (Ser307), five mice per group were randomly selected for injection intraperitoneally with insulin (10 U/kg bw) and sacrificed 10 min after injection. Visceral adipose tissue extracts (eWAT) were immunoblotted with anti-IRS1, anti-phospho-IRS1 (Ser307), anti-AKT1, anti-phospho-Akt (Ser473), anti-GLUT4, antiJNK1, anti-phospho-SAPK/JNK (Thr183/Tyr185), anti-IKKβ, antiTNFα, anti-IL-1β, and internal control anti-β-actin, anti-GAPDH. (IeO) Quantification of proteins described in (H) with normalization to protein levels for each molecule. n = 3 per group. Data are presented as mean ± SEM; *p ≤ 0.05, **p ≤ 0.01 vs CON; †p ≤ 0.05, ††p ≤ 0.01 vs SED; #p ≤ 0.05, ##p ≤ 0.01 HIIT vs MICT by ANOVA; CON, control; HFD, high fat diet; SED, sedentary behavior; HIIT, high intensity interval training; MICT, moderate intensity continuous training; (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

128

Life Sciences 191 (2017) 122–131

N. Wang et al.

Fig. 4. HIIT was particularly prone to enhancing thermogenic activity of brown adipose tissue (BAT) and inducing “browning” of subcutaneous white adipose tissue (scWAT). (A) The gross appearance of interscapular BAT (iBAT) of mice sedentary or trained, showing a striking hypertrophy in the SED mice and an overt deepen of color in the HIIT mice. (B) Representative images of UCP1 immunohistochemistry (IHC) on sections of iBAT isolated from mice sedentary or trained. The iBAT from SED contained more UCP1 + adipocytes than CON group; HIIT and MICT resulted in further emergence of numerous small, multilocular UCP1 + adipocytes, while MICT mice showed weaker change in appearance relative to HIIT. Images are shown at 40× magnification. Scale bars: 200 μm. (C) Quantitative analysis of IHC to determine UCP1 + adipocytes area in iBAT. (D) Representative images of IHC on sections of scWAT isolated from mice sedentary or trained against UCP1. The scWAT from SED contained larger adipocytes than CON group but paralleled UCP1 expression; scWAT from HIIT mice revealed characteristics of beige adipocyte, including the presence of enriched multilocular cells and increased UCP1 expression, but which were not present in the scWAT from MICT mice. Images are shown at 40 × magnification. Scale bars: 200 μm. (E and F) Quantification of average lipid droplet size and UCP1 + adipocytes area of scWAT from mice of all groups. (G) Protein expressions of UCP1, PRDM16 and internal control β-actin in scWAT. (H and I) Quantification of proteins described in (G) with normalization to β-actin level. (J) Transmission electron micrographs of adipocytes from scWAT. Adipocytes of SED mice exhibited several ultrastructural defects: devoid of recognizable organelles, fragmented nuclei, precious few mitochondria, while well-preserved organelles were found in adipocytes from training mice (arrows point to the mitochondria). Images are shown at 15000 ×, selected regions are shown the dense regions of organelles. Scale bar: 0.5 μm. L, lipid droplet; N, nucleus. (K) Quantification of mitochondria number. Data are presented as mean ± SEM, n = 3 per group; *p ≤ 0.05, **p ≤ 0.01 vs CON; †p ≤ 0.05, ††p ≤ 0.01 vs SED; #p ≤ 0.05, ##p ≤ 0.01 HIIT vs MICT by ANOVA; CON, control; HFD, high fat diet; SED, sedentary behavior; HIIT, high intensity interval training; MICT, moderate intensity continuous training. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

of mice following exercise interventions, as a matter of fact that the levels were overall very low in this depot (Fig. S2B). As we know, tissue activity correlates closely with mitochondrial level. Accordingly, HIIT group showed the highest value of mitochondrial population density, actually adipocytes from SED mice contained a limited volume of

determine whether the effects of exercise training on WAT were specific for location, similar experiments were performed with vWAT. Neither HIIT nor MICT showed any impact on UCP1 expression in eWAT in situ, except for the reduced adipocyte size after HIIT (Fig. S2A). Protein expressions of UCP1 and PRDM16 showed blunted induction in vWAT 129

Life Sciences 191 (2017) 122–131

N. Wang et al.

training studies aiming at skeletal muscle [42,43]. It is possible that both the HIIT and MICT exercise failed to affect IKKβ expression in adipose tissue. However, the effect of HIIT or MICT on the IKKβ/NFκB pathway should not be completely excluded and detection of markers for this pathway needs further investigation. Sriwijitkamol et al. [42] demonstrated that training restored abnormalities in IκB and NFκB p50 content without effect on IKKβ phosphorylation or protein content, thus we speculate that training-induced adaptions in IκB/NF-κB signaling may result from activation of other protein kinase rather than IKKβ. These data illustrate that, improvement in insulin sensitivity, along with inflammatory status after HIIT, partly associated with the reduction in JNK but not IKKβ activity in the adipose tissue. Consistently, classical BAT and subcutaneous adipose tissue in rodents have been demonstrated to undergo structural and functional alterations in response to HFD and exercise training. Here, we provide novel evidence that thermogenesis potential is synergistically regulated under conditions of HIIT and HFD, in classical brown fat depots and subcutaneous fat pads. Classical BAT plays an important role in dietinduced thermogenesis and regulation of whole-body energy homeostasis. Therefore, we compared the UCP1 protein, which specifically expressed in BAT mitochondria and largely responsible for dissipation of energy as heat [44]. In concert with previous study [45], training further enhanced UCP1 expression of iBAT in animals with energy surplus diet, in particular HIIT produced stronger stimulus. In addition, scWAT is notable because it can easily “brown” when animals are stimulated with cold, β-adrenergic agonists, or other hormone-like stimuli [46]. We detected distinct phenotypical features of beige adipocytes occurred in scWAT of mice after HIIT, including multilocular adipocytes with decreased lipid volume, and enriched UCP1 + cell populations and mitochondria content. But it did not occur in scWAT of mice after MICT. Further molecular investigation in PRDM16 protein expression is coupled with morphology adaption, which is required for the “browning” of white adipose tissue [47]. Despite we haven't investigated if they have a distinct molecular signature expression from mature brown or white adipocytes, including Tbx1, Tmem26, and Cd137 [48], it is suggested that HIIT program may be better suited to stimulate the activity of BAT and recruitment of beige adipocyte in scWAT. Nevertheless, vWAT seems to be resistant to stimuli that induce “browning” switch. The current study emphasizes that exercise-intensity was an important factor in attenuating HFD-induced adiposity, inflammation, glucolipid metabolic disorders, and white to brown fat phenotypic switch, accounting for the lean phenotype and health-related parameters. Clearly, HIIT is more efficient and time-saving than MICT. As a descriptive study, there are several limitations, especially lacking of evidence from animals with specific gene knockout to clarify whether the effect of exercise training on inflammatory and insulin signal pathway is direct or cascade connected. Additionally, metabolic benefits from “browning” of scWAT and activation of BAT are far more than thermogenesis increment, in future studies, it is important to examine glucose, fatty acid and other mechanism alternations, elucidating the broader spectrum health benefits of fat phenotypic switch to obesity and related comorbidities.

cytoplasm and precious few mitochondria, as evidenced by TEM (Fig. 4J & K). Together, these results indicate that HIIT is more prone to increasing beige cell formation in scWAT but not vWAT. 8. Discussion We have compared here the relative benefits of HIIT for the prevention of obesity and related metabolic disturbances in DIO mice. In this study, two training groups had a similar caloric intake but HIIT protocol exhibited prominent beneficial effects in weight and fat mass loss, adipocyte size decrease, suggesting either a reduced efficiency in energy storage or an increased energy expenditure rate, or both, without inducing a state of energy deficit such as anorexia nervosa found in young athletes following HIIT [33]. What calls for attention is the BW rebound after 6 weeks of training, presumably longer duration and further study is warranted. Consistent with previous studies [7], HIIT is more effective than MICT in treatment of hyperlipidemia induced by HFD. To our knowledge, no studies have compared the relative benefits of HIIT and MICT at the transcriptional level by lipid metabolism genes of the liver, which is an important organ for energy metabolism. The finding of our study was that HIIT restored impaired gene expressions induced by HFD feeding, including that involved in β-oxidation, PPARα, CPT1a and HAD, and genes of molecules related to lipogenesis, SREBP1, ACC1, and FAS. Inferior to HIIT, MICT did not repair the oxidation-related gene disorders. RT-PCR analysis indicates the potential benefits of HIIT on hepatic lipidosis induced by HFD, principally through reverting oxidation-associated gene expressions. Moreover, protein expression of SREBP1 suggests an inhibitory effect of HIIT on the lipogenesis pathway. SREBP1 is a membrane-bound transcription factor that transcriptionally activates genes required for lipogenesis. Among the genes regulated by SREBP1 are ACC, SCD1 and FAS that act as rate-limiting enzymes in the lipid biosynthesis and storage in liver [34,35]. PPARα up-regulates genes involved in cellular fatty acid uptake, transport and oxidation, e.g. CPT1 and HAD [36]. Meanwhile, morphological characteristics confirmed the molecular findings, in that higher content and normal configuration of mitochondria driven by HIIT increase rates of fatty acid oxidation, ultimately giving rise to overall decreased lipid deposition in liver. It is well established that nutrient overload in obesity causes activation of adipocyte stress signaling, in turn, mediates the inflammatory response [37]. The macrophages recruited in adipose tissue are major source of pro-inflammatory cytokines and once activated can further propagate the inflammatory state and interfere with insulin sensitivity. Thus, for obese individuals complicated with IR or diabetes, strategies to reduce the aberrant activation of inflammatory signaling of adipose tissue are of great interest to improve insulin action. Exercise training has been proposed to exert anti-inflammatory effects [38,39], but similar evidence from HIIT is limited. In our study, HOMA-IR index and ITTs indicated HIIT program was more effective at restoring IR in DIO mice. In contrast to the i.p. ITT, the oral GTT did not show significant improvements with HIIT. It is likely related to the route of glucose administration. In male ICR mice, glucose administered by oral gavage caused a weaker increase in blood glucose levels when compared with the same dose of i.p. glucose administration [40]. Circulating and adipose tissue inflammation were mitigated in mice submitted to HIIT, in line with alternations in whole-body and tissue-specific insulin sensitivity enhancement. These changes brought out by HIIT aroused our attention to mechanisms responsible for the synergic effect. It has been known that the substrate for insulin receptor IRS-1 is a tyrosine kinase, and serine phosphorylation of IRS-1, particularly mediated by JNK, reduces insulin receptor signaling [41]; besides, IKKβ/IκB/NF-κB pathway may play an important role in the pathogenesis of IR [42]. In our report, increased JNK phosphorylation in adipose tissue of obese mice declined after HIIT and MICT. However, IKKβ expression did not alter after either HIIT or MICT, inconsistent with previous exercise

Acknowledgments The professional guidance and expert technical assistance from Professor Xiance Sun is gratefully acknowledged.

Grants This research project was funded by Shenyang Science and Technology Plan Projects (F14-231-1-57), and Liaoning Distinguished Professor (Liao taught (2013) No.204). 130

Life Sciences 191 (2017) 122–131

N. Wang et al.

[22] K. Marcinko, S.R. Sikkema, M.C. Samaan, B.E. Kemp, M.D. Fullerton, G.R. Steinberg, High intensity interval training improves liver and adipose tissue insulin sensitivity, Molecular Metabolism 4 (2015) 903–915. [23] P. Bostrom, J. Wu, M.P. Jedrychowski, A. Korde, L. Ye, J.C. Lo, et al., A PGC1-alphadependent myokine that drives brown-fat-like development of white fat and thermogenesis, Nature 481 (2012) 463–468. [24] P.C. Dinas, A. Nikaki, A.Z. Jamurtas, V. Prassopoulos, R. Efthymiadou, Y. Koutedakis, et al., Association between habitual physical activity and brown adipose tissue activity in individuals undergoing PET-CT scan, Clin. Endocrinol. 82 (2015) 147–154. [25] K.I. Stanford, R.J. Middelbeek, K.L. Townsend, M.Y. Lee, H. Takahashi, K. So, et al., A novel role for subcutaneous adipose tissue in exercise-induced improvements in glucose homeostasis, Diabetes 64 (2015) 2002–2014. [26] P. Lee, M.M. Swarbrick, K.K. Ho, Brown adipose tissue in adult humans: a metabolic renaissance, Endocr. Rev. 34 (2013) 413–438. [27] S. Kajimura, B.M. Spiegelman, P. Seale, Brown and Beige fat: physiological roles beyond heat generation, Cell Metab. 22 (2015) 546–559. [28] M. Chondronikola, E. Volpi, E. Borsheim, C. Porter, M.K. Saraf, P. Annamalai, et al., Brown adipose tissue activation is linked to distinct systemic effects on lipid metabolism in humans, Cell Metab. 23 (2016) 1200–1206. [29] O.J. Kemi, J.P. Loennechen, U. Wisloff, O. Ellingsen, Intensity-controlled treadmill running in mice: cardiac and skeletal muscle hypertrophy, J. Appl. Physiol. 93 (2002) 1301–1309. [30] R. Singh, Y. Xiang, Y. Wang, K. Baikati, A.M. Cuervo, Y.K. Luu, et al., Autophagy regulates adipose mass and differentiation in mice, J. Clin. Invest. 119 (2009) 3329–3339. [31] V. Aguirre, T. Uchida, L. Yenush, R. Davis, M.F. White, The c-Jun NH(2)-terminal kinase promotes insulin resistance during association with insulin receptor substrate-1 and phosphorylation of Ser(307), J. Biol. Chem. 275 (2000) 9047–9054. [32] J. Nedergaard, B. Cannon, The changed metabolic world with human brown adipose tissue: therapeutic visions, Cell Metab. 11 (2010) 268–272. [33] V. Singhal, G.D. Maffazioli, K.E. Ackerman, H. Lee, E.F. Elia, R. Woolley, et al., Effect of chronic athletic activity on brown fat in young women, PLoS One 11 (2016) e0156353. [34] P. Cohen, M. Miyazaki, N.D. Socci, A. Hagge-Greenberg, W. Liedtke, A.A. Soukas, et al., Role for stearoyl-CoA desaturase-1 in leptin-mediated weight loss, Science 297 (2002) 240–243. [35] H. Sampath, J.M. Ntambi, Stearoyl-coenzyme A desaturase 1, sterol regulatory element binding protein-1c and peroxisome proliferator-activated receptor-alpha: independent and interactive roles in the regulation of lipid metabolism, Curr. Opin. Clin. Nutr. Metab. Care 9 (2006) 84–88. [36] S. Mandard, M. Muller, S. Kersten, Peroxisome proliferator-activated receptor alpha target genes, Cell. Mol. Life Sci. 61 (2004) 393–416. [37] S. Cinti, G. Mitchell, G. Barbatelli, I. Murano, E. Ceresi, E. Faloia, et al., Adipocyte death defines macrophage localization and function in adipose tissue of obese mice and humans, J. Lipid Res. 46 (2005) 2347–2355. [38] R.L. Bradley, J.Y. Jeon, F.F. Liu, E. Maratos-Flier, Voluntary exercise improves insulin sensitivity and adipose tissue inflammation in diet-induced obese mice, Am. J. Physiol. Endocrinol. Metab. 295 (2008) E586–94. [39] J.M. Bruun, J.W. Helge, B. Richelsen, B. Stallknecht, Diet and exercise reduce lowgrade inflammation and macrophage infiltration in adipose tissue but not in skeletal muscle in severely obese subjects, Am. J. Physiol. Endocrinol. Metab. 290 (2006) E961–7. [40] J.E. Bowe, Z.J. Franklin, A.C. Hauge-Evans, A.J. King, S.J. Persaud, P.M. Jones, Metabolic phenotyping guidelines: assessing glucose homeostasis in rodent models, J. Endocrinol. 222 (2014) G13–25. [41] J. Hirosumi, G. Tuncman, L. Chang, C.Z. Gorgun, K.T. Uysal, K. Maeda, et al., A central role for JNK in obesity and insulin resistance, Nature 420 (2002) 333–336. [42] A. Sriwijitkamol, C. Christ-Roberts, R. Berria, P. Eagan, T. Pratipanawatr, R.A. DeFronzo, et al., Reduced skeletal muscle inhibitor of kappaB beta content is associated with insulin resistance in subjects with type 2 diabetes: reversal by exercise training, Diabetes 55 (2006) 760–767. [43] S. Schenk, J.F. Horowitz, Acute exercise increases triglyceride synthesis in skeletal muscle and prevents fatty acid-induced insulin resistance, J. Clin. Invest. 117 (2007) 1690–1698. [44] D. Ricquier, Respiration uncoupling and metabolism in the control of energy expenditure, Proc. Nutr. Soc. 64 (2005) 47–52. [45] X. Xu, Z. Ying, M. Cai, Z. Xu, Y. Li, S.Y. Jiang, et al., Exercise ameliorates high-fat diet-induced metabolic and vascular dysfunction, and increases adipocyte progenitor cell population in brown adipose tissue, Am. J. Physiol. Regul. Integr. Comp. Physiol. 300 (2011) R1115–25. [46] A. Vitali, I. Murano, M.C. Zingaretti, A. Frontini, D. Ricquier, S. Cinti, The adipose organ of obesity-prone C57BL/6J mice is composed of mixed white and brown adipocytes, J. Lipid Res. 53 (2012) 619–629. [47] P. Cohen, J.D. Levy, Y. Zhang, A. Frontini, D.P. Kolodin, K.J. Svensson, et al., Ablation of PRDM16 and beige adipose causes metabolic dysfunction and a subcutaneous to visceral fat switch, Cell 156 (2014) 304–316. [48] J. Wu, P. Bostrom, L.M. Sparks, L. Ye, J.H. Choi, A.H. Giang, et al., Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human, Cell 150 (2012) 366–376.

Disclosures Authors do not have any conflict of interest to declare. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.lfs.2017.08.023. References [1] P.G. Kopelman, Obesity as a medical problem, Nature 404 (2000) 635–643. [2] D.W. Haslam, W.P. James, Obesity, Lancet 366 (2005) 1197–1209. [3] P. Poirier, T.D. Giles, G.A. Bray, Y. Hong, J.S. Stern, F.X. Pi-Sunyer, et al., Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss: an update of the 1997 American Heart Association scientific statement on obesity and heart disease from the obesity Committee of the Council on nutrition, physical activity, and metabolism, Circulation 113 (2006) 898–918. [4] J.E. Donnelly, S.N. Blair, J.M. Jakicic, M.M. Manore, J.W. Rankin, B.K. Smith, et al., American College of Sports Medicine position stand. Appropriate physical activity intervention strategies for weight loss and prevention of weight regain for adults, Med. Sci. Sports Exerc. 41 (2009) 459–471. [5] G.E. Duggan, D.S. Hittel, C.W. Sensen, A.M. Weljie, H.J. Vogel, J. Shearer, Metabolomic response to exercise training in lean and diet-induced obese mice, J. Appl. Physiol. 110 (2011) 1311–1318. [6] M. Rakobowchuk, S. Tanguay, K.A. Burgomaster, K.R. Howarth, M.J. Gibala, M.J. MacDonald, Sprint interval and traditional endurance training induce similar improvements in peripheral arterial stiffness and flow-mediated dilation in healthy humans, Am. J. Physiol. Regul. Integr. Comp. Physiol. 295 (2008) R236–42. [7] A.E. Tjonna, S.J. Lee, O. Rognmo, Stolen TO, A. Bye, P.M. Haram, et al., Aerobic interval training versus continuous moderate exercise as a treatment for the metabolic syndrome: a pilot study, Circulation 118 (2008) 346–354. [8] A.E. Tjonna, Stolen TO, A. Bye, M. Volden, S.A. Slordahl, R. Odegard, et al., Aerobic interval training reduces cardiovascular risk factors more than a multitreatment approach in overweight adolescents, Clin. Sci. 116 (2009) 317–326. [9] L.J. Whyte, J.M. Gill, A.J. Cathcart, Effect of 2 weeks of sprint interval training on health-related outcomes in sedentary overweight/obese men, Metab. Clin. Exp. 59 (2010) 1421–1428. [10] U. Wisloff, A. Stoylen, J.P. Loennechen, M. Bruvold, O. Rognmo, P.M. Haram, et al., Superior cardiovascular effect of aerobic interval training versus moderate continuous training in heart failure patients: a randomized study, Circulation 115 (2007) 3086–3094. [11] G.A. Gaesser, S.S. Angadi, High-intensity interval training for health and fitness: can less be more? J. Appl. Physiol. 111 (2011) 1540–1541. [12] M. Buchheit, P.B. Laursen, High-intensity interval training, solutions to the programming puzzle. Part II: anaerobic energy, neuromuscular load and practical applications, Sports Med. 43 (2013) 927–954. [13] M. Cocks, C.S. Shaw, S.O. Shepherd, J.P. Fisher, A. Ranasinghe, T.A. Barker, et al., Sprint interval and moderate-intensity continuous training have equal benefits on aerobic capacity, insulin sensitivity, muscle capillarisation and endothelial eNOS/ NAD(P)Hoxidase protein ratio in obese men, J. Physiol. 594 (2016) 2307–2321. [14] A.C. Corte de Araujo, H. Roschel, A.R. Picanco, D.M. do Prado, S.M. Villares, A.L. de Sa Pinto, et al., Similar health benefits of endurance and high-intensity interval training in obese children, PLoS One 7 (2012) e42747. [15] M.J. Gibala, J.P. Little, M.J. Macdonald, J.A. Hawley, Physiological adaptations to low-volume, high-intensity interval training in health and disease, J. Physiol. 590 (2012) 1077–1084. [16] C. Frosig, A.J. Rose, J.T. Treebak, B. Kiens, E.A. Richter, J.F. Wojtaszewski, Effects of endurance exercise training on insulin signaling in human skeletal muscle: interactions at the level of phosphatidylinositol 3-kinase, Akt, and AS160, Diabetes 56 (2007) 2093–2102. [17] J.P. Little, J.B. Gillen, M.E. Percival, A. Safdar, M.A. Tarnopolsky, Z. Punthakee, et al., Low-volume high-intensity interval training reduces hyperglycemia and increases muscle mitochondrial capacity in patients with type 2 diabetes, J. Appl. Physiol. 111 (2011) 1554–1560. [18] J.P. Little, A. Safdar, G.P. Wilkin, M.A. Tarnopolsky, M.J. Gibala, A practical model of low-volume high-intensity interval training induces mitochondrial biogenesis in human skeletal muscle: potential mechanisms, J. Physiol. 588 (2010) 1011–1022. [19] N.A. Johnson, T. Sachinwalla, D.W. Walton, K. Smith, A. Armstrong, M.W. Thompson, et al., Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss, Hepatology 50 (2009) 1105–1112. [20] A.B. Jordy, M.J. Kraakman, T. Gardner, E. Estevez, H.L. Kammoun, J.M. Weir, et al., Analysis of the liver lipidome reveals insights into the protective effect of exercise on high-fat diet-induced hepatosteatosis in mice, Am. J. Physiol. Endocrinol. Metab. 308 (2015) E778–91. [21] T.T. Tran, C.R. Kahn, Transplantation of adipose tissue and stem cells: role in metabolism and disease, Nat. Rev. Endocrinol. 6 (2010) 195–213.

131