Lyophilized Maqui (Aristotelia chilensis) Berry Induces Browning in the Subcutaneous White Adipose Tissue and Ameliorates the Insulin Resistance in High Fat Diet-Induced Obese Mice

Maqui (Aristotelia Chilensis) berry features a unique profile of anthocyanidins that includes high amounts of delphinidin-3-O-sambubioside-5-O-glucoside and delphinidin-3-O-sambubioside and has shown positive effects on fasting glucose and insulin levels in humans and murine models of type 2 diabetes and obesity. The molecular mechanisms underlying the impact of maqui on the onset and development of the obese phenotype and insulin resistance was investigated in high fat diet-induced obese mice supplemented with a lyophilized maqui berry. Maqui-dietary supplemented animals showed better insulin response and decreased weight gain but also a differential expression of genes involved in de novo lipogenesis, fatty acid oxidation, multilocular lipid droplet formation and thermogenesis in subcutaneous white adipose tissue (scWAT). These changes correlated with an increased expression of the carbohydrate response element binding protein b (Chrebpb), the sterol regulatory binding protein 1c (Srebp1c) and Cellular repressor of adenovirus early region 1A–stimulated genes 1 (Creg1) and an improvement in the fibroblast growth factor 21 (FGF21) signaling. Our evidence suggests that maqui dietary supplementation activates the induction of fuel storage and thermogenesis characteristic of a brown-like phenotype in scWAT and counteracts the unhealthy metabolic impact of an HFD. This induction constitutes a putative strategy to prevent/treat diet-induced obesity and its associated comorbidities.


Introduction
Stimulation of the brown adipose tissue (BAT) and the induction of browning in white adipose tissue (WAT) as a strategy against obesity and its associated metabolic complications has generated growing interest in recent years. This interest is based on the ability of both BAT and browned-WAT to increase energy expenditure (EE) mainly through fatty acid consumption [1,2], and their pivotal role in the control of energy homeostasis in mammals [3,4]. In addition to classical brown adipocytes located in BAT, thermogenic adipocytes with similar characteristics can be found within white adipose tissue (WAT) [3]. These brite/beige adipocytes are metabolically and phenotypically similar to brown adipocytes and can actively contribute to increasing whole-body EE. Specifically, brite/beige adipocytes show a multilocular phenotype and express genes closely related to BAT metabolism (Ucp1 as a marker of its thermogenic capacity in addition to genes implied in de novo lipogenesis (DNL), fatty acid oxidation (FAO), lipolysis, etc.).
Recent evidence shows that the activation of BAT and the induction of browning in WAT can be induced by cold acclimation but also by nutritional inputs under different signaling cascades [2,[5][6][7]. Cold is a classic activator of BAT and of beige adipocyte development and function [5,8]. Regarding nutritional inputs, we recently demonstrated that low-protein diets and the cooked-tomato sauce called "sofrito" are able to induce Ucp1 expression in WAT, thus indicating a browning phenotype [6,7]. In the same way, other authors published that high-fat diets, bioactive compounds and prebiotics can also induce browning in WAT [9][10][11][12][13][14].
Part of the cold-induced metabolic profile in BAT is regulated by the stimulation of carbohydrate-responsive element binding protein b (ChREBPb) through the Ak strain transforming/protein kinase 2 (AKT2) activity [15]. Besides ChREBP, fibroblast growth factor 21 (FGF21) has shown beneficial effects on glucose/lipid homeostasis and body weight control among other mechanisms by increasing energy expenditure (EE) and inducing browning and UCP1 overexpression in adipose tissues [6,[16][17][18][19], as well as by promoting the insulin-dependent glucose uptake, mitochondrial biogenesis, and adiponectin secretion in adipocytes [20,21]. In this case, it has been widely demonstrated that FGF21 activity and/or signaling respond to nutritional challenges [22].
With the global aim of deepening the knowledge of the molecular mechanisms responsible for the metabolic effects of maqui and in some way of the anthocyanidin-rich foods, we investigated the effect of a lyophilized maqui berry preparation on the onset and progression of the diet-induced obesity (DIO) in mice subjected to high-fat diet (HFD) for 16 weeks. We studied the impact of maqui in the metabolic profile of the obese phenotype. We focused on the adipose tissue metabolic phenotype because of its role on the progression of obesity and as a major target to counteract the onset and development of this pathology and its metabolic-associated diseases such as insulin resistance. We analyzed specifically the subcutaneous WAT (scWAT) because growing evidence suggests that this depot is protective to metabolic health while visceral is detrimental [38][39][40][41][42][43].
Globally, our results highlight the potential role of maqui in the treatment of diet-induced obesity and insulin resistance. Further studies will be needed to identify the effects of maqui on healthy population and the impact of its regular consumption as part of a healthy dietary pattern such as the Mediterranean diet.

Anthocyanins Determinations by UPLC-DAD
For sample preparation, 0.1 g of maqui was extracted with 5 mL of a mixture of water and ethanol 80:20 (v/v). The extraction was repeated 3 times to increase extractability of the anthocyanidins. The extract was evaporated under vacuum to remove the ethanol and reconstituted to a final volume of 10 mL with MilliQ water. This procedure was done in triplicate. Finally, the sample was filtered through a 0.22 µm PTFE membrane filter into an amber vial for UPLC analysis.
The quantification and identification of anthocyanins was done using a Waters Acquity Ultra Performance Liquid Chromatography H class (Waters Corp, Milford MA, USA) coupled to Photodiode Array (PDA) detector (Waters Corp, Milford MA, USA). The chromatographic separation was performed by an Aquity BEH C18 column 2.1 mm × 100 mm, 1.7 µm. The chromatographic method proposed by Andrés-Lacueva et al. (2005) adapted to the UPLC system was used [44]. Briefly, the injection volume was 10 µL and the gradient elution was performed with water/5% formic acid (v/v) (A) and 100% acetonitrile (B) at a constant flow rate of 0.75 mL/min. A decreasing linear gradient of solvent A was used. Separation was carried out in 11 min under the following conditions: 0 min, 98% A; 6 min, 95% A; 9 min, 90% A; 9.1 min, 20% A; 9.3min, 20% A; 9.4 min, 92% A, 11 min, 92% A.

Animal Procedures-Dosage Regimen
Animal procedures were approved by the Animal Ethics Committee of the University of Barcelona (CEEA-173/18). C57BL/6J littermates' male mice (n = 23) were housed in a temperature-controlled room (22 ± 1 • C) on a 12-h/12-h light/dark cycle and were provided free access to commercial rodent chow and tap water prior to the experiments. When animals were four-week-old it was confirmed that all animals were normoglycemic before being randomly assigned into two experimental groups (HFD and HFD supplemented with maqui (HFDM)). Both groups were fed a diet of 45% fat-derived calories (HFD) (D12451, Research Diets) for 16 weeks supplemented or not with lyophilized maqui. HFD group (n = 9) had free access to HFD diet and filtered-tap water and HFDM (n = 14) group had free access to HFD diet and filtered-tap water supplemented with maqui (20 mg of lyophilized maqui/mL of filtered-tap water). To prepare the supplemented water, 1 g of the lyophilized maqui was added to 50 mL of tap-filtered water. This mixture was prepared extemporaneously every two days to prevent the oxidation of maqui bioactive compounds.
The dosage regimen of maqui was calculated according to the polyphenol intake recommended as beneficial by the Predimed Study (820 mg in a human diet of 2300 kcal) [45,46]. Mice intake is around 10-15 kcal per day, which means 4-5 mg of polyphenols per day scaling-down the recommended beneficial quantity in humans. Table 1 shown that 1g of lyophilized maqui provides 45 mg of anthocyanins. The dose of maqui was adjusted to achieve 4 mg of polyphenols per day, considering that mice take 2-3 mL of water/day. The nutritional composition of the lyophilized maqui used (Maquiberry, Native for Life, Chile) is indicated in the supplemental information (Table S1).
During the 16-week nutritional intervention, food and beverage intake were recorded every two days and body weight twice a week. At the end of the nutritional intervention, the animals were euthanized. Blood was extracted by intracardiac puncture, and serum was obtained by centrifugation (1500 rpm, 20 min). Epididymal WAT (eWAT), scWAT and BAT were isolated, immediately snap-frozen and stored at −80 • C for future analysis.

RNA Isolation and Quantitative RT-PCR
Total RNA was isolated from frozen tissues using TRI Reagent™ Solution (AM9738, Thermo Fisher Scientific, Waltham, USA), followed by DNaseI treatment (K2981, Thermo Fisher Scientific, Waltham, USA). cDNA was synthesized from 1 µg of total RNA using the High-Capacity cDNA Reverse Transcription Kit (4368814, Thermo Fisher Scientific, Waltham, USA). Relative mRNA levels were measured by quantitative PCR (qPCR) using SYBR™ Select Master Mix for CFX (4472942, Thermo Fisher Scientific, Waltham, USA). 18S and B2M were used as housekeeping genes. The sequences of the primers used in qPCR are presented in Table S2. Results were obtained by the relative standard curve method, and values were referred to the HFD group.

Glucose Tolerance Test (GTT) and Insulin Tolerance Test (ITT)
Mice were fasted for 6 h in the morning, and then injected intraperitoneally (i.p.) with 1.5 g of glucose (Sigma)/kg mouse (GTT) or 0.5 UI of insulin solution (Sigma)/kg mouse (ITT). Blood samples were collected from the tail vein, and glucose levels were measured using a glucometer (Glucocard SM, Menarini, Florence, Italy) prior to the i.p. injection and at 30, 60 and 120 min postinjection. GTT was performed 14 weeks after the beginning of maqui supplementation and ITT on week 15th.

Histological Analysis
For the histological analysis, pieces of scWAT of each animal were fixed in 10% formalin (Sigma) and embedded in paraffin. Afterwards, 4 µm-thick sections were cut and stained with hematoxylin and eosin (H&E). Images were acquired using a Digital Upright Microscope BA310 Digital and a Moticam 2500 camera. The selection of the test objects was performed according to color and choosing the same limits for binarization for all images. At least three pictures from different regions of each cut were taken.

Data Analysis/Statistics
GraphPad Prism version 8.02 (GraphPad, San Diego, CA, USA) was used to perform the statistical analyses. Two tailed Student's Test with Welch's correction when not equal SDs can be assumed was used to determine significant differences among experimental groups. The statistical analysis of body weight progression, body weight/calorie intake and curves of GTT and ITT was performed by 2 factor ANOVA with pot-hoc test (Bonferroni's). In all cases, p-value < 0.05 was considered statistically significant. All data are expressed as the mean ± SEM.

Maqui Dietary Supplementation Reduces HFD-Induced Body Weight Gain and Improves Insulin Sensitivity in Mice
C57BL6/J mice fed an HFD for 16 weeks put on weight (Figure 1a,b) and displayed glucose intolerance (Figure 1e,f) and insulin resistance (Figure 1g,h). Comparing both experimental groups revealed that mice fed HFD supplemented with maqui (HFDM) showed an attenuated progression in body weight (Figure 1a-c), even the animals were hyperphagic, and their daily caloric intake was higher ( Figure 1d). Concretely, the HFDM animals put on less weight after the 16 weeks of nutritional intervention ( Figure 1b). Even though no statistical differences were observed in the body weight progression (Figure 1a), there were when the ratio between body weight and caloric intake was calculated ( Figure 1c). These data indicated that, in some way, there is an increased energy expenditure in these animals.
Regarding the insulin/glucose responsiveness, HFDM mice shown a significant reduction in fasting glucose at Week 15 of the nutritional intervention (207.6 ± 5.5 vs. 166.5 ± 6.1, p = 2.7 × 10 −5 ). It is worth highlighting that, after the glucose injection, the HFDM animals displayed a better glucose curve ( Figure 1e) that corresponded to a significant reduction of the area under the curve (AUC) of the GTT (Figure 1f, p < 0.05). In the case of the ITT, the curve of glucose after the insulin injection showed significantly lower levels of glucose at the first time-points ( Figure 1g). These differences were minimized but maintained over time, even though they did not reach statistical significance ( Figure 1g). Finally, the AUC of the ITT showed a clear tendency to a reduction in HFDM ( Figure 1h, p < 0.08). Although some of the data are not significant individually, altogether they indicate an improvement on glucose tolerance and insulin sensitivity after 16 weeks of maqui dietary supplementation.   16-week nutritional intervention with maqui. Body weight was recorded twice a week. The graph represents the mean ± SEM of weekly increments in both experimental groups. (b) Total body weight increment in grams after 16-week nutritional intervention with maqui. The graph represents the mean ± SEM of the total body weight increment in both experimental groups. (c) Body weight (g) related to calorie intake (kcal) per week for the 16-week nutritional intervention with maqui. Calorie intake was calculated based on the energy density of the HFD and the amount of food consumed daily. In HFDM mice, the kcal from maqui were also added. The graph represents the mean ± SEM of the weekly body weight increase related to calorie intake in both experimental groups.  16-week nutritional intervention with maqui. Body weight was recorded twice a week. The graph represents the mean ± SEM of weekly increments in both experimental groups. (b) Total body weight increment in grams after 16-week nutritional intervention with maqui. The graph represents the mean ± SEM of the total body weight increment in both experimental groups. (c) Body weight (g) related to calorie intake (kcal) per week for the 16-week nutritional intervention with maqui. Calorie intake was calculated based on the energy density of the HFD and the amount of food consumed daily. In HFDM mice, the kcal from maqui were also added. The graph represents the mean ± SEM of the weekly body weight increase related to calorie intake in both experimental groups.

Maqui Dietary Supplementation Induces a Multilocular Phenotype in scWAT
HFDM mice were leaner than their littermates (Figures 1a and 2a) and had more BAT and less WAT (Figure 2b,c). Because of the healthier profile observed in HFDM mice regarding body weight and insulin resistance, we hypothesized that maqui could be exerting its effects through the induction of browning in scWAT. The H&E staining of scWAT revealed that maqui supplementation induced the transition of unilocular adipocytes to multilocular ones (Figure 2d). No differences in other WAT depots due to maqui supplementation were observed in any analysis performed.
HFDM mice were leaner than their littermates (Figures 1a and 2a) and had more BAT and less WAT (Figure 2b,c). Because of the healthier profile observed in HFDM mice regarding body weight and insulin resistance, we hypothesized that maqui could be exerting its effects through the induction of browning in scWAT. The H&E staining of scWAT revealed that maqui supplementation induced the transition of unilocular adipocytes to multilocular ones (Figure 2d). No differences in other WAT depots due to maqui supplementation were observed in any analysis performed. In the context of browning, recent publications have shown the relevance of fat-specific protein 27 (FSP27) isoforms in the phenotype of unilocular or multilocular lipid droplets. FSP27 is considered to be, at least in part, the protein responsible for the formation and growth of lipid droplets (LDs). Two isoforms have been described with different expression patterns and functions. Fsp27a is expressed in WAT, where it promotes the formation of large LDs. By contrast, BAT expresses the Fsp27b isoform that contributes to the ensemble of smaller LDs [53,54]. To confirm the brown-like In the context of browning, recent publications have shown the relevance of fat-specific protein 27 (FSP27) isoforms in the phenotype of unilocular or multilocular lipid droplets. FSP27 is considered to be, at least in part, the protein responsible for the formation and growth of lipid droplets (LDs). Two isoforms have been described with different expression patterns and functions. Fsp27a is expressed in WAT, where it promotes the formation of large LDs. By contrast, BAT expresses the Fsp27b isoform that contributes to the ensemble of smaller LDs [53,54]. To confirm the brown-like phenotype of scWAT in HFDM mice, the mRNA levels of both Fsp27 isoforms were measured. The results indicated that the dietary maqui supplementation causes a shift in the expression pattern of Fsp27. The scWAT depot of HFDM mice showed a significant reduction in the levels of Fsp27a and a significant induction of Fsp27b expression (Figure 2e), thus reinforcing the idea that maqui causes a brown-like phenotype in the scWAT.
phenotype of scWAT in HFDM mice, the mRNA levels of both Fsp27 isoforms were measured. The results indicated that the dietary maqui supplementation causes a shift in the expression pattern of Fsp27. The scWAT depot of HFDM mice showed a significant reduction in the levels of Fsp27a and a significant induction of Fsp27b expression (Figure 2e), thus reinforcing the idea that maqui causes a brown-like phenotype in the scWAT.

Maqui Induces the Expression of Chrebpa, Chrebpb and Glut4 in scWAT
As mentioned above, part of the cold-induced metabolic profile in BAT is regulated by the stimulation of ChREBPb through the AKT2 activity [15]. In WAT, the expression of Chrebp in adipose tissue is regulated by GLUT4-mediated glucose uptake that activates the ChREBPa isoform, which in turn induces the expression of Chrebpb, an alternative-promoter transcribed isoform. Chrebpb expression in human adipose tissue predicts insulin sensitivity, and its induction has been highlighted as an effective strategy for preventing and treating obesity-related metabolic dysfunction and type 2 diabetes [55][56][57]. Globally, ChREBP is considered the major regulator of DNL in adipose tissue [58]. To analyze the putative role of ChREBPb and GLUT4 in the induction of the metabolic changes observed in scWAT due to maqui supplementation, the mRNA levels of Glut4, Chrebpa and Chrebpb were evaluated. The data show that maqui increases the expression of these three genes (Figure 4), thus providing insight into the molecular mechanism through which maqui induces a brown-like phenotype in scWAT. that are considered the control group, which produces an arbitrary value of 1. Data are presented as the mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 versus the HFD group.

Maqui Induces the Expression of Chrebpa, Chrebpb and Glut4 in scWAT
As mentioned above, part of the cold-induced metabolic profile in BAT is regulated by the stimulation of ChREBPb through the AKT2 activity [15]. In WAT, the expression of Chrebp in adipose tissue is regulated by GLUT4-mediated glucose uptake that activates the ChREBPa isoform, which in turn induces the expression of Chrebpb, an alternative-promoter transcribed isoform. Chrebpb expression in human adipose tissue predicts insulin sensitivity, and its induction has been highlighted as an effective strategy for preventing and treating obesity-related metabolic dysfunction and type 2 diabetes [55][56][57]. Globally, ChREBP is considered the major regulator of DNL in adipose tissue [58]. To analyze the putative role of ChREBPb and GLUT4 in the induction of the metabolic changes observed in scWAT due to maqui supplementation, the mRNA levels of Glut4, Chrebpa and Chrebpb were evaluated. The data show that maqui increases the expression of these three genes (Figure 4), thus providing insight into the molecular mechanism through which maqui induces a brown-like phenotype in scWAT. Bars represent the fold induction in the mRNA levels versus the HFD animals that are considered the control group and assigned an arbitrary value of 1. Data are presented as the mean ± SEM. **p < 0.01, ***p < 0.001 versus the HFD group.

Maqui Increases the Expression of Adiponectin and FGF21 and the FGF21 Signaling in the scWAT.
Obesity is considered an FGF21-resistant state usually due to a downregulation of the fibroblast growth factor receptor (FGFR) and/or its co-receptor β-klotho (KLB) that impairs FGF21 signaling [59,60]. To analyze the effect of maqui berry supplementation on the FGF21 signaling, the mRNA levels of Fgf21 itself, Fgfr1, Fgfr4 and KLB were determined. Moreover, to analyze the functionality of the FGF21 signal transduction pathway, the expression levels of Egr-1 and adiponectin were measured. Figure 5 shows that, in scWAT, both the FGF21 and its receptors (FGFR1 and FGFR4) were significantly overexpressed in HFDM mice versus the HFD mice ( Figure 5). These data, together with the induction of Egr-1 and adiponectin in scWAT ( Figure 5), indicate that maqui supplementation improves FGF21 signaling and effectiveness in scWAT. Bars represent the fold induction in the mRNA levels versus the HFD animals that are considered the control group and assigned an arbitrary value of 1. Data are presented as the mean ± SEM. ** p < 0.01, *** p < 0.001 versus the HFD group.

Maqui Increases the Expression of Adiponectin and FGF21 and the FGF21 Signaling in the scWAT
Obesity is considered an FGF21-resistant state usually due to a downregulation of the fibroblast growth factor receptor (FGFR) and/or its co-receptor β-klotho (KLB) that impairs FGF21 signaling [59,60]. To analyze the effect of maqui berry supplementation on the FGF21 signaling, the mRNA levels of Fgf21 itself, Fgfr1, Fgfr4 and KLB were determined. Moreover, to analyze the functionality of the FGF21 signal transduction pathway, the expression levels of Egr-1 and adiponectin were measured. Figure 5 shows that, in scWAT, both the FGF21 and its receptors (FGFR1 and FGFR4) were significantly overexpressed in HFDM mice versus the HFD mice ( Figure 5). These data, together with the induction of Egr-1 and adiponectin in scWAT ( Figure 5), indicate that maqui supplementation improves FGF21 signaling and effectiveness in scWAT.

Discussion
Globally, our data show that maqui dietary supplementation ameliorates the unhealthy effects of HFD. By using diet-induced obese mice, the present study demonstrated that supplemented animals displayed better insulin responsiveness, decreased weight gain and increased thermogenic activity. The analysis of gene expression in different tissues showed that scWAT exhibited differential expression of genes involved in browning, DNL, mitochondrial and peroxisomal FAO, multilocular adipocytes and thermogenesis. These changes were probably related with the increased expression of Chrebpa, Chrebpb, Glut4, Creg1 and Srebp1c but also the improvement in FGF21signaling.
Obesity is essentially caused by an imbalance between energy intake and energy expenditure. Some evidence indicates that, at some point, the white adipose tissue (WAT) fails to adequately keep the surplus of nutrients, which together with an insufficient differentiation of new adipocytes leads to an off-WAT accumulation of ectopic lipids in peripheral relevant organs that may cause the metabolic obesity-related metabolic dysfunctions. It seems obvious that defects in WAT functionality together with peripheral lipotoxicity are key in the onset of metabolic syndrome [61][62][63][64][65][66].
The biomedical relevance of brown and beige adipocytes lies in the ability of these cells to increase EE and counteract metabolic diseases such as obesity or type 2 diabetes [1,2]. Indeed, increasing the activity of brown, beige or both fat depots is considered a promising strategy for the treatment of metabolic diseases [67,68].
Our results clearly confirm the previously described effects of maqui regarding glucose metabolism but also describe the capacity of maqui to induce browning in HFD-induced obese mice. Dietary-supplemented obese mice showed less weight gain despite their hyperphagic behavior, thus indicating that in some way these animals have a higher metabolic rate, probably due to an increased BAT depot and a more energy consuming scWAT. The absence of effects on other WAT depots indicated that scWAT was the target for maqui beneficial effects against diet-induced obesity and

Discussion
Globally, our data show that maqui dietary supplementation ameliorates the unhealthy effects of HFD. By using diet-induced obese mice, the present study demonstrated that supplemented animals displayed better insulin responsiveness, decreased weight gain and increased thermogenic activity. The analysis of gene expression in different tissues showed that scWAT exhibited differential expression of genes involved in browning, DNL, mitochondrial and peroxisomal FAO, multilocular adipocytes and thermogenesis. These changes were probably related with the increased expression of Chrebpa, Chrebpb, Glut4, Creg1 and Srebp1c but also the improvement in FGF21signaling.
Obesity is essentially caused by an imbalance between energy intake and energy expenditure. Some evidence indicates that, at some point, the white adipose tissue (WAT) fails to adequately keep the surplus of nutrients, which together with an insufficient differentiation of new adipocytes leads to an off-WAT accumulation of ectopic lipids in peripheral relevant organs that may cause the metabolic obesity-related metabolic dysfunctions. It seems obvious that defects in WAT functionality together with peripheral lipotoxicity are key in the onset of metabolic syndrome [61][62][63][64][65][66].
The biomedical relevance of brown and beige adipocytes lies in the ability of these cells to increase EE and counteract metabolic diseases such as obesity or type 2 diabetes [1,2]. Indeed, increasing the activity of brown, beige or both fat depots is considered a promising strategy for the treatment of metabolic diseases [67,68].
Our results clearly confirm the previously described effects of maqui regarding glucose metabolism but also describe the capacity of maqui to induce browning in HFD-induced obese mice. Dietary-supplemented obese mice showed less weight gain despite their hyperphagic behavior, thus indicating that in some way these animals have a higher metabolic rate, probably due to an increased BAT depot and a more energy consuming scWAT. The absence of effects on other WAT depots indicated that scWAT was the target for maqui beneficial effects against diet-induced obesity and that at least part of the effects of berries on glucose metabolism and insulin sensitivity go through the improvement of adipose tissue functionality.
BAT thermogenesis is stimulated by adrenergic induction of cAMP production, which activates protein kinase A (PKA) to drive transcription of thermogenic genes (including Ucp1), lipolysis and FAO. Although less recognized, the activation of BAT thermogenesis paradoxically induces the anabolic DNL pathway [15,[82][83][84]. Several studies support the important role of ChREBP isoforms in the control of insulin signaling and lipogenesis in adipose tissue and describe the induction of both isoforms' activity under the Glut4-dependent glucose uptake [55,[85][86][87][88]. In WAT, HFD feeding lowers the expression of Chrebp and DNL genes in mice [55]. Moreover, the expression of a constitutively active ChREBP in WAT protects mice against obesity and insulin resistance by among others reducing adiposity and increasing the expression of gene related to adipocyte differentiation and browning [89]. In the same context, an adipose-specific Chrebp knockout mice show a decrease in DNL and are insulin resistant with an impaired insulin action in the liver, muscle and fat [85]. Beyond mice, in humans, the expression of Chrebp and lipogenic genes in WAT shows a strong correlation with insulin sensitivity, and the improvement of insulin sensitivity in insulin-resistant people restores Chrebp and glucose transporter type 4 (Glut4) expression in adipose tissue [56].
Our key finding is that this characteristic metabolic feature of BAT also appears in the scWAT of obese mice fed a maqui-supplemented HFD where there is an induction of Chrebpa, Chrebpb and Glut4 expression and also higher mRNA levels of genes involved in DNL, multilocular LDs formation and thermogenesis/browning. The remaining question is how ChREBP is activated under maqui supplementation. The increased levels of Glut4 in HFDM mice indicated that glucose or glucose metabolites could be the major inducers of Chrebp expression [90,91]. Moreover, although ChREBP was initially identified as a glucose-responsive factor, recent evidence suggests that it is also essential for fructose-induced lipogenesis both in the small intestine and liver [55,92]. The effects of maqui could, therefore, be attributed to its fructose content. However, the absence of Chrebpb induction in liver (data not shown) allows us to rule out this possibility.
SREBP1c is also a key regulator of hepatic DNL [93][94][95][96][97][98]. It has been demonstrated that, in the liver, SREBP1c and ChREBP are both necessary for the expression of lipogenic and glycolytic genes [88,99] In adipose tissue, the role of SREPB1c is more controversial. In adipose tissue, the mRNA levels of lipogenic genes did not change in animals using a Srebp1c loss of function or gain of function approach, thus indicating that in this tissues Srebp1c is not essential for DNL activation [58]. By contrast, mice under caloric restriction (CR) showed an SREBP1c/PGC1a-dependant induction of DNL in adipose tissue giving to Srebp1c an role on activating this metabolic pathway under this specific nutritional condition [100]. Finally, CREG1 has been described as an inducer of Ucp1 and FGF21 expression in an adipocyte P2-Creg1-transgenic (Tg) mice and globally of BAT adipogenesis and browning [101,102].
Our results indicate that maqui supplementation is able to induce the expression of Chrebpb, Srebp1c, Pgc1a and Creg1, thus we cannot discard any of them as possible contributors to the induction of DNL and thermogenic genes observed.
In addition, in the scWAT of dietary-supplemented obese mice, there was an increase in the expression of peroxisomal FAO enzymes that would make possible the contradiction of having FAO and DNL simultaneously active. Peroxisomal FAO is important in this scenario where, despite an increased expression of Cpt1b, the rate-limiting enzyme of mitochondrial FAO, this pathway would be inhibited by the malonyl-CoA produced by DNL.
Apart from the abovementioned mechanisms our data also pointed out the improvement of FGF21 signaling as a way through which maqui could exert its beneficial effects. It has been widely described that FGF21 levels are increased in obesity and diabetes in both animal models and humans [60,103,104]. The downregulation of FGF21 receptors in adipose tissue seems to be the key point to explain the FGF21-resistant state described mainly in obese mice as well as, in some studies, in humans [43,59,60,[105][106][107][108]. In this context, the restoring of FGF21 signaling can be considered as a potential therapeutic strategy to improve the metabolic parameters of obese individuals and to reduce the risk of obesity-related diseases and some evidence support this hypothesis [7,109]. In various rodent models of diet-induced obesity a positive correlation between the beneficial effects of polyphenol-rich fruit extracts and FGF21 has been described. This correlation with FGF21 can be due to an induction of FGF21 levels [110,111] or an improvement of the FGF21 signaling [7,112,113]. Finally, FGF21-resitance in adipose tissue has been linked to a decreased production of adiponectin [114]. Adiponectin induces fatty acid oxidation leading to a reduction of ectopic lipids and finally the improvement of insulin sensitivity [115]. In our results, the induction of the mRNA levels of adiponectin, FgfR1, FgfR4 and Egr1 in scWAT of HFDM mice indicates that, in obesity, maqui supplementation increases the sensitivity to FGF21 of scWAT.
Although our results do not allow us to discard that other signaling pathways could stimulate the white to beige/brown transition described, the improvement of FGF21 signaling together with the overexpression of ChREBP, CREG1, PGC1a and SREBP1c are at least key players in the induction of browning described under maqui supplementation. Neuroendocrine signaling or molecules such as leptin or Bmp8b will be analyzed in further studies to try to complete the signaling cascade activated by maqui [2,67,116].
To summarize, we demonstrated that a nutritional intervention with maqui partially alleviates the unhealthy effects of HFD in mice. Our results provide evidence that, in mice, a dietary supplementation with maqui added to beverage activates the induction of fuel storage and thermogenesis characteristic of a brown-like phenotype in scWAT. Finally, based on previously published data, our results indicate that maqui could exert its effects, at least in part, through the induction of Chrebpb expression and the improvement of FGF21 signaling. Finally, it is worth mentioning that in this work the dose of maqui was scaled-down from the polyphenol intake recommended as beneficial in humans by the Predimed Study [45,46]. This is important because several phenolic compounds such as resveratrol, quercetin, cyanidin-3-glucoside (C 3 G), capsaicin, hesperidin have green tea extract have been described as inducers of BAT activity or WAT browning but in most cases the dietary supplementation was performed using high doses and only the active compounds [9][10][11][12][13][117][118][119][120][121].

Limitations of the Present Work and Further Studies
This work clearly demonstrates that maqui is effective in obese mice. The impact of maqui in healthy individual needs to be evaluated. We do not have data about the effects of maqui on normal chow-fed animals where neither obesity nor insulin resistant is present, but this experimental approach will be included in our further studies. Positive results in this follow-up experiment will allow pointing out the efficacy of the consumption of maqui in the prevention of some metabolic diseases and whether to include its regular consumption as part of a heathy dietary pattern.
On the other hand, despite the observed changes in gene expression, together with the scWAT and BAT appearance and the histological analysis to define properly the browning phenotype, in this study, the translation to protein was assumed. To overcome this limitation, Western blot analyses will be performed to reinforce our results and deepen the mechanism of action of maqui.

Conclusions
In conclusion, our data provide evidence that, in obese mice, a dietary intervention with a regular dose of an anthocyanidin-enriched berry (maqui) can induce a browning phenotype in scWAT and improve partially the insulin sensitivity, thus ameliorating some of the unhealthy effects of HFD. These effects reinforce the anthocyanidin-enriched foods as a potential strategy to prevent or treat type 2 diabetes and obesity-related diseases and point out maqui as a putative functional fruit to counteract at least in part obesity and its metabolic complications. The data presented in this manuscript reinforce the inclusion of maqui in the diet of obese individuals.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2076-3921/8/9/360/s1, Table S1: Maqui extract nutritional composition, Table S2: Sequences of the primers used in SYBR Green assays and references of the probes used in Taqman assays, Figure S1: Funding: This study was supported by the Ministerio de Economía y Competitividad (grants AGL2017-82417-R to PFM and DH and AGL2016-75329-R to RML-R), by the Generalitat (grants 2017SGR683 and 2017SGR196), by the Associació Catalana de la Diabetis to JR (grant 600004-Ajut ACD a la recerca en diabetis 2017), Instituto de Salud Carlos III, ISCIII from the Ministerio de Ciencia, Innovación y Universidades (CIBEROBN-AEI/FEDER, UE). VS was supported by Conicyt's fellowship from the Government of Chile. UM-G was supported by Conacyt's fellowship from the Government of México. JMC was supported by BEC.AR fellowship from Ministerio de Educación, Cultura, Ciencia y Tecnología, Argentina. The APC was funded by the University of Barcelona.