A Novel Multi-Ingredient Supplement Activates a Browning Program in White Adipose Tissue and Mitigates Weight Gain in High-Fat Diet-Fed Mice

Ethics approval was granted by the McMaster University Animal Research Ethics Board and conformed to the standards of the Canadian Council on Animal Care (AUP#16-05-15). Male C57/BL6J diet-induced obesity mice were ordered from Jackson Laboratories (Bar Harbor, ME, USA) and placed on a high-fat diet (HFD) containing 60% energy from fat at 6 weeks of age, with the exception of chow (CHOW) control animals. All animals (total n = 178) were provided with food and water ad libitum and were housed in separate standard microisolator cages with enrichment (12 h light/dark cycle at 22 °C) to accurately record food consumption.

At ~12 weeks of age, mice were allocated into experimental groups which were standardized by body weight (BW). For ~1 month, the experimental groups were fed 1 of the following HFDs containing ~60% energy from fat: HFD control (HFD), HFD containing mitochondria-enhancing agents, HFD containing weight loss agents, or HFD containing a combination of both mitochondria and weight loss-enhancing agents or a standard chow diet. All diets are listed in Supplementary Table S1. In brief, mitochondria-enhancing agents included beetroot extract, coenzyme Q10 (CoQ10), alpha lipoic acid (α-LA), vitamin E, and creatine [21,22,23,24,25,28,33]. Weight loss agents included green tea extract (GTE), black tea extract (BTE), green coffee bean extract (GCBE), conjugated linoleic acid (CLA), and forskolin [27,29,30,34,35]. To demonstrate the effects of a healthy and/or standard diet on body composition in sedentary conditions, a group of mice were fed chow diet throughout (CHOW). To evaluate the effects of exercise (EX) and combination therapy (HFD with nutritional supplementation and exercise), animals were housed in cages containing an exercise wheel and underwent structured exercise (3 days/week).

To determine which dietary composition and/or intervention was able to reduce BW and fat mass (FM) most effectively, we examined animals (n = 10/group, standardized pre-intervention by BW) provided with either the control or experimental diet, with or without exercise. Prior to the introduction of the experimental diets, baseline testing was conducted on all animals. Baseline testing included the following measurements: BW, relative FM and lean mass (LM), muscle strength, and maximal running capacity. During the intervention, food consumption and BW were measured. All anthropometric, organ weight, food consumption, and exercise performance data for this preliminary experiment are listed in the appended supplemental data (Figure S1A–H). We determined that the composition comprised of both mitochondria-enhancing agents and weight loss agents (ME10) provided the most robust changes in body composition and weight, the effect of which was enhanced with exercise (ME10 + EX). These changes were more robust than when a composition of mitochondrial-enhancing agents or, separately, a composition of weight loss agents were provided (Figure S1A) to the animals. These results led to the next experimental approach (see below).

To verify that the dietary composition was able to: (i) effectively reduce BW and FM, (ii) enhance exercise-mediated improvements, (iii) evaluate an ME with fewer ingredients (ME7), and (iv) evaluate potential biological mechanism(s) for the observed weight loss, we repeated the experiment. We examined an independent cohort of animals (n = 12/group, standardized by BW). The control HFD, CHOW, and ME10 diets were identical to those of Study 1, as were the exercise groups (EX, ME10 + EX). To define the minimal, essential components of the ME10 supplement driving the observed weight loss in the preliminary study, we eliminated BTE, CLA, and creatine, and termed this 7-component supplement ME7. The ME7 group received HFD containing only 3 of the weight loss agents (forskolin, GTE, GCBE) and only 4 of the mitochondria-enhancing agents (BE, CoQ10, Vit E, α-LA), removing creatine, BTE, and CLA (Supplementary Table S1).

To evaluate the effects of ME7 and ME10 supplementation on fat oxidation and basal energy expenditure compared to HFD control following a short-term period of supplementation, we performed metabolic monitoring using the Comprehensive Lab Animal Monitoring System (CLAMS; Columbus Instruments), with lipid oxidation rates collected via indirect calorimetry and activity levels measured in beam-breaks across the x-axis of the metabolic cages similar as previously described [36]. C57/BL6 (n = 24) diet-induced obesity mice were ordered from Jackson Laboratories and provided the same HFD (HFD) at 6 weeks of age. At ~12 weeks of age, mice were allocated into groups (n = 8/group) standardized by BW, placed in the metabolic cages, and allowed to acclimate to the new environment for ~12 h. Following the acclimation period, mice remained on the HFD control diet for a period of 1 day (baseline; BSL) to collect metabolic data for each group (prior to any change in diet). After 1 day of measurements (BSL) on the HFD control, mice were administered HFD, HFD + ME7, or HFD + ME10 diets for a 3-day period. Diet was removed from the animals prior (as they slept) to sacrifice on the third day of experimental diet provision.

At the onset of the study and at an age of 6 weeks, 65 g of the designated diet was provided to each cage. During the experimental period the feed weight was measured, ensuring that feed was removed from both the hamper and cage bedding. Food consumption was calculated as the difference in total feed weight over 24 h. When the feed reached 25 g, an additional 40 g of feed was added to ensure there was always sufficient food.

Relative FM and LM were quantified using a time-domain nuclear magnetic resonance (NMR) whole-body composition analyzer (Minispec LF90II, Bruker; Billerica, MA, USA) and normalized to body weight.

Animal exercise capacity was measured at baseline and endpoint via endurance stress test performed on a treadmill. All animals were placed in individual treadmill lanes and allowed to acclimate for 5 min prior to the initiation of the endurance stress test. The exercise session was preceded by a 10-min warm-up at 10 m/min. Following this period, the speed was increased by 1 m/min until animals had reached exhaustion. Exhaustion was determined at the point when the animals remained unresponsive to hind limb stimulation via paintbrush for >5 s. The speed and time of exhaustion of each individual animal was recorded.

To address the mitigation of obesity through exercise [37], endurance exercise training was performed by select groups (EX, ME10 + EX). These groups performed structured endurance exercise on an Exer 6 M treadmill (Columbus Instruments, Columbus, OH, USA) for 45 min/session, 3 days/week (alternate days) for the duration of the study to complement nightly cage wheel running activity. Exercise sessions were preceded by the same 5-min acclimation and 10-min warm-up at 10 m/min as the endurance stress test. After initial familiarization, animals were run at a speed of 15 m/min. Animals were encouraged to run on the treadmill with gentle hind-limb stimulation via paintbrush when they stopped running. All animals completed all exercise bouts. The animals in these groups were independently housed in voluntary running wheel cages and utilized the running wheels freely.

Mice were briefly anesthetized with isoflurane (Abraxis Bioscience, Los Angeles, CA, USA). Animals were euthanized via exsanguination, with blood being collected for downstream analysis. The tibialis anterior (TA) muscles, the quadriceps (left quadricep; SkM), liver, inter-abdominal epididymal white adipose tissue (eWAT), and brown adipose tissue (BAT) were all weighed, harvested, and snap-frozen using liquid nitrogen. The right quadricep was prepared for high-resolution respirometry. Snap-frozen samples were then placed in −80 °C freezer for long-term storage. For Study 3, tissues were harvested under isoflurane sedation to maximize the detection of the phosphorylation status of AMP-activated protein kinase (AMPK) [38].

Muscle fibers were permeabilized similarly to previously described [39]. High-resolution O₂ consumption measurements were conducted in MiR05 buffer using the Oroboros Oxygraph-2k (Oroboros Instruments, Corp., Innsbruck, Austria). In brief, we examined ADP-stimulated respiration, with pyruvate and malate added as complex I substrates (via generation of NADH to saturate electron entry into complex I), followed by a titration of submaximal ADP and maximal ADP. To examine complex I + II state 3 respiration, this was followed by addition of succinate (PMDG + S). Complex II kinetics were separately determined through the stepwise addition of succinate in the presence of ADP and rotenone (to inhibit and prevent superoxide generation at complex I). The detailed methods can be found in the Supplementary Experimental Procedures.

Tissues were prepared and homogenized in Lysing Maxtrix D tubes (MP Biomedicals, Solon, OH, USA) via the FastPrep-24 Tissue and Cell Homogenizer (MP Biomedicals) in ice-cold RIPA lysis and extraction buffer (ThermoFisher Scientific, Waltham, MA, USA, 89901) supplemented with protease and phosphatase inhibitor cocktail (Halt^TM, ThermoFisher Scientific, 78440). Protein concentrations were then determined using a BCA Protein Assay Kit (Pierce, Thermo Scientific, 23225). Samples were subsequently prepared in Laemmli Buffer (Fisher Scientific, AAJ61337AC). Immediately prior to electrophoresis, samples were heated at 95 °C for 5 min and centrifuged, unless they were used for mitochondrial complex expression analyses. These samples were heated to 37 °C, followed by centrifugation. Samples were then loaded on a 4–20% Criterion TGX Precast Midi Protein Gel (Bio-Rad Laboratories, Hercules, CA, USA 5671094) with a PageRuler Plus Prestained Protein Ladder Protein Ladder at 10 to 250 kDa (Thermo Fisher, 26619), and run in 1X Tris/Glycine/SDS buffer (Bio-Rad Laboratories, 1610732). Following electrophoresis, samples were transferred using Trans-Blot Turbo Midi Nitrocellulose Transfer Packs (Bio-Rad Laboratories, 5671094) and stained with Ponceau S solution (Sigma, P7170-1L) for 5 min. The membranes were imaged and placed in blocking solution (5% BSA in 1X TBS) for 1 h, prior to overnight primary antibody incubation at 4 °C. All primary antibodies were prepared in 5% BSA in TBS. The antibodies were the OXPHOS (oxidative phosphorylation) cocktail, SOD1, SOD2, 4-HNE, UCP1, AMPK total protein, and phosphorylated AMPK (Thr172), and were used at a concentration of 1:1000 (information in Supplementary Table S2). The membranes were washed in 1X TBST and incubated with the appropriate secondary antibody (Jackson Immunoresearch 715-035-151, 715-035-152) at 1:10,000 (5% BSA in 1X TBS) for 1 h. The membranes were washed again in 1X TBST, then incubated with Clarity Western ECL substrate (Bio-Rad Laboratories, 170-5061) containing 50% luminol/enhancer and 50% peroxide solution for 5 min. Membranes were developed using ChemiDoc MP Imaging System (Bio-Rad Laboratories). Band intensities were quantified using Image Lab 6.1 software (Bio-Rad Laboratories) and normalized to Ponceau S staining.

RNA was extracted from tissue, following lysis in Trizol Reagent (Life Technologies, Burlington, ON, Canada), into Lysing Maxtrix D tubes (MP Biomedicals, Solon, OH, USA) using the FastPrep-24 Tissue and Cell Homogenizer (MP Biomedicals). The RNA phase was subsequently purified using the PureLink™ RNA Mini Kit (Thermo Scientific, cat. no. 12183025) as per the manufacturer’s instructions. RNA concentrations (ng⋅mL⁻¹) and purity (260/280) were determined with the Nano-Drop 1000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Samples were then reverse-transcribed using a high-capacity cDNA reverse transcription kit (SuperScript^® VILO™ Master Mix; Invitrogen, Waltham, MA, USA, cat. no. 11755050).

The absolute abundance for specific mRNA transcripts was assayed using TaqMan^® Fast Advanced Master Mix (cat. no. 4444963) and TaqMan Gene Expression Assays purchased from ThermoFisher Scientific (Supplementary Table S3). This was performed in a CFX384 Fast Real-Time PCR System (Bio-Rad Laboratories). Relative gene expression was calculated using the comparative method as described [40]. Briefly, C_t values were normalized to the housekeeping gene peptidylprolyl isomerase A (Ppia), which was unchanged between groups and has previously been shown to be a stable housekeeping reference gene [41].

A Milliplex MAP Mouse High Sensitivity T Cell Panel (Millipore, Billerica, MA, USA) was performed on a Luminex 200 (Luminex, Austin, TX, USA) system using Xponent 3.1 software (Luminex), as per the manufacturer’s instructions. Quantification was performed by the Analytical Facility for Bioactive Molecules of the Centre for the Study of Complex Childhood Diseases, The Hospital for Sick Children, Toronto, Canada.

The maximal activities of citrate synthase (CS), cytochrome c oxidase (COX), and short chain β-hydroxy-acyl-CoA-dehydrogenase (β-HAD) were determined in homogenized SkM, as previously described [42]. All enzyme assays were performed utilizing a calibrated spectrophotometer (Cary Bio-300, Varion, Inc., Palo Alto, CA, USA). Detailed methods found in Supplementary Experimental Procedures.

Briefly, tissues were immersed in 10% formalin with 90% PBS for 24 h and placed into 70% EtOH for paraffin embedding and subsequent hematoxylin & eosin (H&E) staining by the John Mayberry Histology facility at McMaster University Medical Centre. All imaging was undertaken on a Nikon 90i microscope under 10X magnification, and adipocyte size analysis was performed with the NIS Elements Analysis Software (Nikon, Inc., Melville, NY, USA).

Statistical analysis was performed by GraphPad Prism 7.00 (San Diego, CA, USA). To examine differences between groups, 1-way ANOVA was employed. If omnibus F-test(s) were significant, a Fisher’s least squares difference (LSD) post hoc test was used to determine group differences compared to only the HFD. For pre- and post-supplementation data (e.g., body weight, body fat percentage) a 2-way repeated measures ANOVA (group × time) with post-hoc used when appropriate. For respiration data, a 2-way ANOVA (titration x group) with post hoc was used where appropriate. Statistical significance was accepted at p < 0.05. All results are presented as means ± SEM.

By design, there were no significant differences in total body mass in the ME10, ME7, EX, or ME10 + EX groups compared to the HFD control mice prior to the intervention. After 4 weeks of intervention the HFD control animals showed a 31.2 ± 1.2% increase in BW (p < 0.05, Figure 1B). BW was unchanged in the ME10 (+3.9 ± 2.1%, p > 0.05) and ME7 (−3.0 ± 1.1%, p > 0.05) groups; both groups were significantly lower than those of the HFD group (p < 0.05, Figure 1B) following the intervention. The EX (−4.0 ± 2.4%) and ME10 + EX (−12.7 ± 2.1%) groups had a significant reduction in BW; both groups had significantly lower BW as compared to the HFD group (p < 0.05, Figure 1B). The sedentary CHOW group did not change significantly in weight (+2.2 ± 0.7%, Figure 1B) during the experiment.

We examined whether the marked weight loss following ME supplementation was global or tissue specific. HFD control animals had a greater liver weight, consistent with hepatomegaly (p < 0.05, Figure 1C). Both with and in the absence of exercise, animals given ME10 and ME7 had larger relative TA weights compared to HFD, respectively (p < 0.05, Figure 1D). Conversely, all animals on experimental diets and/or performing exercise interventions had a lower relative WAT weight (p < 0.05, Figure 1E). There were no differences in relative BAT weight between the HFD and supplementation diet/intervention groups, but CHOW animals did have a lower BAT depot size (p < 0.05, Figure 1F). Fat and muscle organ weights confirmed findings in vivo via NMR. We observed significant increases in FM (%) in the HFD control group, whereas animals supplemented with ME7 and ME10 were unchanged. In addition, the EX and ME10 + EX groups showed a significant reduction following intervention (p < 0.05, Figure 1G). Following the intervention, all groups had a significantly lower body fat percentage than the HFD group. The HFD and sedentary CHOW animals had a significant decrease in relative lean mass (LM) over the intervention period. Lean mass was unchanged with ME7 and ME10 but tended to increase in EX (p = 0.07) and increased in combinatory ME10 + EX (p < 0.05, Figure 1H). All groups had a significantly higher relative lean mass as compared to the HFD animals following the intervention.

Throughout the intervention period, relative food intake (g/BW/day) was significantly decreased in HFD animals but increased in the experimental diet groups during the invention (p < 0.05, Figure S2A). Absolute food intake (g/day) was lower in ME7 and ME10 animals (p < 0.05, Figure S2B) compared to HFD, but remained unchanged in the groups in which the greatest mean weight loss occurred (EX, ME10 + EX, p > 0.05).

To evaluate the effects of ME7 and ME10 supplementation independently or in conjunction with exercise training (ME10 + EX) upon aerobic capacity, an exercise capacity test was performed. Following the intervention, the HFD control and ME10 group exhibited significant decreases in maximal running capacity, whereas the ME7, EX, and ME10 groups did not (p < 0.05, Figure S2C).

Mitochondria OXPHOS-related protein content in eWAT improved in a complex-specific manner in response to supplementation and/or exercise. Compared to HFD, supplemented and/or exercising animals had greater complex I, II, and IV content (p < 0.05, Figure 2A). The final common pathway indicating browning/‘beiging’ is the induction of UCP1 protein expression and mitochondrial biogenesis [15]. UCP1 protein was detectable in eWAT, although considerable variation was observed within groups. The ME7 group exhibited the greatest mean expression compared to the HFD group (p > 0.05, Figure 2A). We observed a similar pattern in Ucp1 mRNA expression (p > 0.05, Figure 2B), as well as greater mRNA expression of markers of the browning program in eWAT, including Cidea, Prdm16, Pppara, and Adipoq, in experimental diet compositions (ME7, ME10) and exercise groups (p < 0.05, Figure 2B).

Oxidative stress may play a role in the initiation or exacerbation of obesity [10,43]. Exercise has been shown to have antioxidant effects in muscle [44,45]. Therefore, we examined whether supplementation and/or chronic exercise facilitates antioxidant defense during a continued HFD challenge. We found a significantly greater SOD1 protein content in the ME10, EX, and combinatory groups relative to HFD (p < 0.05, Figure 2C), with no observable differences in SOD2 (Figure 2C) or lipid peroxidation products (4-HNE) (Figure 2C).

Adipose tissue plays an important role in the regulation of lipid availability [46], and thus we examined whether markers of lipid transport and β-oxidation were improved with supplementation. We observed higher mRNA expression of markers responsible for cellular (Fatp1) and mitochondrial (Cpt2) uptake of fatty acids in experimental groups compared to HFD (p < 0.05, Figure 2E), with non-significant but similar trends in cytosolic membrane transport (Cpt1b). This dysregulation was mirrored by mRNA expression of enzymes linked to mitochondrial fatty acid β-oxidation (Hadh, Lcad) (p < 0.05, Figure 2D).

Brown adipose tissue (BAT) is important for thermogenesis and energy homeostasis, has a high mitochondrial content, and plays a role in the development of diet-induced obesity [47]. Paradoxically, the BAT response to exercise has shown counter-intuitive effects, with a reduction in BAT mass and UCP1 content [48]. We observed that in the HFD + ME7 group, there was significantly higher complex I expression as compared with the HFD control group (p < 0.05, Figure 2E). However, there was a significantly lower expression in complex IV protein in ME7 (p < 0.05) as well as complex III and IV in the EX and ME10 + EX groups as compared to the HFD group in BAT (p < 0.05, Figure 2E). There was no significant difference in UCP1 expression between experimental groups (p > 0.05). We observed higher mRNA abundance related to mitochondrial function, lipid fusion, ROS defense, and energy homeostasis in BAT primarily in the EX and ME10 + EX groups (p < 0.05, Figure 2F).

To confirm the influence of HFD + ME7 supplementation on mitochondrial OXPHOS protein expression in epidydimal WAT, we followed up with an examination of inguinal white adipose tissue (ingWAT) in a subset of animals (HFD, n = 6; HFD + ME7, n = 6). These findings showed a consistent and significant reduction in adipocyte size for ingWAT (p < 0.05, Figure 2H,I) as well as eWAT (p < 0.05, Figure 2J,K) in the ME7 group as compared to the HFD control group. Confirming our previous observations in the epidydimal WAT depot, OXHPHOS complexes I, II, and V were significantly greater in ME7 compared to HFD (p < 0.05, Figure 2L). As in eWAT, UCP1 protein was detectable in ingWAT and mean expression was higher in ME7 (p < 0.05, Figure 2L).

The release of inflammatory cytokines from adipocytes [49,50] can cause the recruitment of immune cells to growing WAT, which propagates local and systemic inflammation [51]. Therefore, we examined supplementation and/or exercise on circulating markers of inflammation via a multiplex immune-bead assay. IL-8 (KC), GM-CSF, IL-2, IL-6, IL-7 concentrations were lower in all groups compared to the HFD animals (p < 0.05, Figure 3A), with non-significant but similar trends in IL-1β and TNF-α (p > 0.05). To confirm whether the decrease in systemic cytokines reflected an altered expression in adipose tissue, we measured markers of inflammation in eWAT. Similarly, there was lower mRNA expression in Tnfa, Il1b, Il6, and Casp1 in animals undergoing supplementation and/or exercise (p < 0.05, Figure 3B).

To examine skeletal muscle mitochondrial function, we utilized high-resolution respirometry and complex-specific assays. ADP-stimulated respiration was greater in the SkM of HFD (p < 0.05, Figure 4A) compared to all other groups (main effect for group). Notably, animals in all experimental groups demonstrated ADP-stimulated oxygen consumption rates that were similar in nature to those of the CHOW-fed animals (i.e., suggesting a normalization of mitochondrial function). SkM complex II-supported respiration (succinate-induced FADH₂ synthesis) was significantly higher in the HFD group compared to all experimental groups (p < 0.05, Figure 4B) across physiological, submaximal, and maximal concentrations of succinate (main effect for group). SkM Complex I + II state 3-supported respiration (PMDG + S) was significantly higher in HFD animals compared to all experimental groups (p < 0.05, Figure 4C), with the exception of ME10 (p = 0.08).

Electron transport chain protein content measured in the quadriceps was largely unchanged in the experimental groups. Despite a relatively short exercise period, there was a higher abundance of complex IV protein in the ME10 + EX group as compared to HFD animals (p < 0.05, Figure 4D). These findings suggest the differences in observed via high-resolution respirometry were not due to changes in mitochondrial content. There was no significant intervention effect on SkM β-HAD maximal activity (a marker of β-oxidation capacity), CS maximal activity, mitochondrial COX maximal activity (p > 0.05, Figure 4E), or the ratio of COX/CS (the COX activity relative to total mitochondria; data not shown p > 0.05).

ME10 + EX animals had increased SOD1 SkM protein content (p < 0.05, Figure 4F), while ME10 animals showed a trend toward an increase in SOD1 protein content (p = 0.06, Figure 4F). ME10 + EX mice showed a significantly higher SOD2 content (p < 0.05, Figure 4F). Lipid peroxidation-derived 4-HNE did not significantly differ between experimental interventions (p > 0.05, Figure 4F).

In both 4-week supplementation experimental approaches (Approach 1; Figure S1, and Approach 2; Figure 1), acute weight loss occurred at as early as after the first 3 days of supplementation. Therefore, we examined lipid oxidation rates and activity via CLAMS following a 3-day period (Figure 5A). Lipid oxidation was unchanged in the HFD group but increased in the ME10 and ME7 groups (p < 0.05, Figure 5B), such that values for the ME7 group were significantly higher than those of HFD animals (p < 0.05). Activity levels of ME7 and ME10 mice increased after 3 days of supplementation, whereas the HFD group showed significantly reduced activity (p < 0.05, Figure 5C). Food consumption (g/24 h) during the experimental period was increased in the HFD group but was unchanged in the ME10 and ME7 groups (p < 0.05, Figure 5D). AMPK signaling is indirectly activated by catecholamine secretion in response to β-adrenergic stimulation [52] and has previously been shown to be upregulated in response to some ME components such as GTE [53]; therefore, immunoblots for p-AMPK_Thr172 were performed. ME10 supplementation increased AMPK phosphorylation (p < 0.05, Figure 5E) in eWAT. There were no significant differences detected in SkM (p > 0.05, Figure 5F).