Acute Administration of Calafate (Berberis microphylla) Extract Induces the Expression of Thermogenic Markers and Modulates Gut Microbiota in Mice Fed a High-Fat Chow Diet

Obesity, an abnormal increase in white adipose tissue (WAT) due to an energy imbalance, is a major health problem worldwide [1]. Alongside WAT, brown adipose tissue (BAT) contributes to the regulation of body temperature through thermogenesis [2]. The mitochondria of BAT express uncoupling protein 1 (UCP1), which dissipates energy in the form of heat [3]. Factors such as thyroid hormones, catecholamines, exposure to cold, and dietary phytochemicals can enhance the thermogenic activity of BAT [4]. It has also been reported that some stimuli may also promote the appearance of clusters of brown adipocytes in certain regions of the WAT [5].

Recently, polyphenols have been considered as possible anti-obesity compounds due to their thermogenic characteristics [6]. These secondary plant metabolites are poorly absorbed in the small intestine and pass into the colon where they are deglycosylated by the gut microbiota (GM) and their aglycone moiety subsequently metabolized by different bacterial taxa into a large number of metabolites that can be absorbed [7]. Although some polyphenol glycosides can be hydrolyzed by enzymes present in enterocytes, such as the lactase-phlorizin hydrolase, most of them are hydrolyzed by the GM [8].

The GM forms a complex bacterial ecosystem of almost 100 billion microorganisms. A greater richness of species is observed in the GM of healthy individuals, and the loss of this diversity is known as dysbiosis [9]. GM diversity is influenced by several factors, of which diet is one of the most important [10]. Recent data show that the metabolism of polyphenols by GM influences their bioavailability and bioactivity [11].

Calafate (Berberis microphylla) is a native Patagonian berry rich in anthocyanins, hydroxamic acid derivatives, and flavonols, which has been shown to exert a wide range of health benefits including anti-obesogenic and pro-thermogenic effects in animal models [12, 13]. The aim of this study was to determine the impact of acute treatment with a Calafate polyphenol-rich extract on the thermogenic activity of brown and beige adipose tissue and on the GM composition of mice fed a high-fat diet. The role of GM in this effect was also determined by using a group of animals treated with antibiotics.

Calafate extract derived from freeze-dried Calafate (INIA, Chile) was obtained using a low-toxicity solvent (ethanol:water). The crude extract was filtered and purified by HPLC using an Amberlite XAD7HP resin column. Characterization revealed a total polyphenol concentration of 4,257 mg GAE/100 g, total anthocyanins of 4,028 mg D-3-G/100 g, and a total antioxidant capacity of 59,009 mmol TE/100 g.

Eight-week-old C57BL/6J male mice (n = 30; approximate weight 20 g) from the colony of the Nutrition Department of the University of Chile, were housed in cages of 3–4 animals, at controlled relative humidity, room temperature (21–23°C), with 12-h light-dark cycles, and were randomly assigned into 4 groups: (a) control group (C) fed a standard diet (D12328 diet, 11 kcal% fat, 16% protein, 73% carbohydrate with corn starch Surwit diet) (n = 5); (b) obese group (HF) fed a high-fat diet (D12330 diet, 58 kcal% fat, 16% protein, 26% carbohydrate with corn starch Surwit diet) (n = 10), (c) obese group fed the high-fat diet (D12330 diet) that received 50 mg/kg body weight per day of total polyphenols of Calafate extract in water for 3 weeks from week 21–23 (n = 10) (HFC), (d) obese group fed the high-fat diet (D12330 diet) that received a cocktail of broad-spectrum antibiotics [14] (ampicillin [0.5 mg/mL], neomycin [0.5 mg/mL], gentamicin [0.5 mg/mL], metronidazole [0.5 mg/mL], vancomycin [0.25 mg/mL]) in sterile drinking water ad libitum for 2 weeks (week 19–20) prior to Calafate extract administration, and then treated daily with the Calafate extract for 3 weeks (week 21–23) (n = 5) (HFCAB). The control and the HF diets were administered for 4 months. The dose of Calafate extract administered to the animals in the HFC and HFCAB groups was previously tested in our laboratory and the results published [15]. Antibiotic administration in the HFCAB group was performed to test whether GM was necessary to obtain the thermogenic effects of the Calafate extract. At the end of trial, animals were euthanized, their body weights recorded, and samples of inguinal (iWAT), epididymal (eWAT), and interscapular (BAT) adipose tissue collected, as well as the cecal contents, and stored at −80°C until subsequent analysis

The AllPrep^® DNA/RNA/Protein Mini Kit (Qiagen, Hilden, Germany) was used to obtain RNA, DNA, and proteins from samples of adipose tissue. RNA was treated with DNase (DNA-free, Amboina). 500 ng of total RNA was used for reverse transcription using a commercial kit (High-Capacity cDNA Reverse Transcription Kit, ThermoFisher Scientific, Waltham, MA, USA). The resulting cDNA was used for qPCR using TaqMan^® probes (Applied Biosystems) in a Stratagene Mx3000P System (Agilent Technologies, CA, USA). Specific transcripts of thermogenesis-related markers (Ucp1, Dio2, Pgc1-α, Pparα, Pparγ, Prdm16, and Sirt1) were evaluated in inguinal WAT and interscapular BAT. The expression levels of genes were normalized using Gapdh and Ppia housekeeping genes. The reported gene expression changes between groups were calculated by the 2 (⁻∆∆Ct) method [16].

BAT was fixed in glutaraldehyde (2.5%) and dissected into bundles of fibers. Samples were then washed with water and stained with 1% uranyl acetate for 2 h. Finally, 80-nm sections were cut, mounted, and examined on a Tecnai G2 T12 transmission electron microscope (Philips-FEI). The mitochondrial size was calculated as the mean area of each organelle [17, 18]. Mitochondria cristae parameters were evaluated manually by trained personnel, as previously described [17, 18]. Morphometric analyses were formed by using ImageJ software [17, 18].

Cecal samples were collected for GM analysis. Bacterial DNA was extracted using a commercial kit (QIAamp DNA Stool Mini Kit, Qiagen, Hilden, Germany), adding a bead-beater stage to optimize bacterial breakdown [19]. DNA concentrations (ng/μL) were measured on a NanoQuant Infinite 200Pro kit (Tecan, Männedorf, Switzerland), and DNA integrity was assessed by electrophoresis on 1% agarose gel. Samples were sent to the University of Illinois (Roy J. Carver Biotechnology Center - Urbana-Champaign, Champaign, IL, USA) for PCR amplification of the 16 s rDNA gene (v3–v4 region) and amplicon sequencing using the MiSeq Illumina platform (Illumina, San Diego, CA, USA). Results were expressed as relative abundances of each bacterial taxa (phyla, families, and genera) and the intra- and interindividual diversity indices were also determined.

Statistical analysis was performed using SPSS Packaged 20.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 6.0 software (GraphPad Software, Inc., San Diego, CA, USA) for graphing results. Normal distribution of the data was determined by using the Shapiro-Wilk test. Data were described as mean and standard deviations (for variables with parametric distribution) or as median and interquartile range (for these with nonparametric distribution). Differences between groups were determined by one-way ANOVA test followed by Tukey’s post hoc test in case of parametric data, and Kruskal-Wallis test followed by Dunn’s post hoc in case of nonparametric data. Relative bacterial abundances were expressed as proportions, i.e., percentages of the total sequences. Additionally, principal coordinate analysis with weighted UniFrac [20] was performed to analyze the clustering of samples based on their composition. A value of p < 0.05 was considered statistically significant.

The general characteristics of the animals in the four groups are shown in Table 1. Weight gain throughout the experiment was higher in the HF and HFC groups than in the HFCAB and C groups. The eWAT weight was significantly higher in the groups fed the high-fat diet compared to the C group. BAT weight was significantly higher in the HFCAB group than in the C group (Table 1).

As shown in Figures 1a, b, no differences were found in the mitochondrial size between groups. However, the relative distribution of mitochondrial size (Fig. 1e) showed a greater proportion of larger mitochondria (+1.3 µm²) in the HFC group than in the C, HF, and HFCAB groups. Compared to C group, HF and HFCAB groups had a lower number of mitochondrial cristae (Fig. 1c) and total length of the cristae (Fig. 1d). Animals from the HFC group displayed similar mitochondrial cristae number and total cristae length than in the C group (Fig. 1c, d).

Changes in gene expression of the biomarkers of thermogenesis in BAT and iWAT are described in Table 2. In both tissues, a higher expression of Dio2 transcript was observed in the HFC group compared to the C and HFCAB groups, while no differences were observed with Ucp1, Pgc1-α, Pparϒ, Pparα, Prdm16, and Sirt1. Therefore, antibiotic administration prevented the Calafate-induced increase of Dio2 in BAT and iWAT.

Regarding the GM, Shannon index as a marker of alpha diversity is shown in Figure 2. The HFCAB group displayed a lower alpha diversity than the other groups, confirming the impact of antibiotic administration on GM composition. The principal component analysis (Fig. 3) showed a significant difference in β-diversity. Both the HF and the HFC groups clustered together and differed from the C and HFCAB groups, showing that the antibiotic treatment and high-fat diet administration affect the GM composition.

Changes in the composition of bacterial phyla, families, and genera are described in Table 3. Results show that the high-fat diet, alone or with Calafate extract, had no effect on the microbiota at the phylum level. Animals from the HF group had lower abundances of the Burkholderiaceae, Bacteroidales F082 and Coriobacteriales families, and a higher abundance of Lachnospiraceae, compared to C group. At the genus level, animals fed the high-fat diet had a lower abundance of Parasutterella, Muribaculum,PrevotellaceaeUCG-003, and Rikenellaceae RC9 gut group, and a higher abundance of Tyzzerella 3, Desulfovibrio, and Lachnospiraceae NK4A136 group, versus the C group. Administration with Calafate extract prevented (fully or in part) the increase of Tyzzerella 3, Desulfovibrio, and Lachnospiraceae NK4A136 group, and the decrease of Prevotellaceae UCG-003 and Rikenellaceae RC9 gut group, induced by the high-fat diet. Finally, HFC group also showed a higher abundance of Ruminococcaceae UCG-010 genus versus the HF group.

The administration of antibiotics affects the relative abundance of various bacterial groups in the HFCAB group. This group exhibited a lower abundance of the Proteobacteria and Deferribacteres phyla, and a higher abundance of Verrumicrobia, compared to C group. At the family level, it also had a lower abundance of Burkhohlderiaceae, Bacteroidales F082, Rikenellaceae, Deferribactetaceae, and Prevotellaceae while that of Peptococcaceae and Akkermansiaceae was higher, compared to C group. Moreover, in the HFCAB group, the genera Erysipelatoclostridium, Akkermansia, Lachnospiraceae NK4A136 group, and the species Akkermansia muciniphila were more abundant, while Parasutterella, Mucispirillum, Alistipes, and Rikenellaceae RC9 gut group were less abundant, compared to C group.

In this study, we were able to demonstrate that acute administration of Calafate increased the relative expression of Dio2, a thermogenesis marker, and prevented alterations in mitochondrial cristae and GM composition induced by the consumption of a high-fat diet. Contrarily to previous studies [12, 13] which observed a decreased weight gain after 16 weeks of administration of the same dose of Calafate extract, the current study showed no effect of the administration of Calafate extract for 3 weeks to mice fed a high-fat diet on this parameter. Probably, this lower impact can be attributed to the shorter duration of Calafate extract administration in the present study.

Interestingly, one of the most notable findings of our study is that mitochondria from the HF group presented lower mitochondrial cristae density compared to control. Mitochondrial cristae is a dynamic energetic compartment involved in oxidative phosphorylation and ATP production, whose shape changes according to physiological conditions, as shown in mononuclear cells [17] and skeletal muscle cells [21]. Decrease in mitochondrial cristae density has been associated with oxidative stress, an early risk factor for metabolic alterations in the obesity [22]. Thus, administration of Calafate may have prevented mitochondrial cristae alterations in BAT of obese mice, probably due to the antioxidant properties of its polyphenols.

Type II iodothyronine deiodinase is a protein encoded by the DIO2 gene that is involved in BAT activation and thermogenesis [23‒25]. We observed increased expression of the Dio2 transcript in the HFC group compared to the C group in both BAT and iWAT. This increase was not observed in the HFCAB group, although they also received Calafate. This could be caused by the administration of antibiotics and the consequent decrease in GM alpha diversity in the HFCAB group. Previous studies indicate that GM could stimulate Type II iodothyronine deiodinase activity and expression in BAT and beige adipose tissue through the stimulation of the TGR5 receptor by secondary bile acids formed from primary bile acids by specific bacterial taxa, resulting in increased thermogenesis and browning [26].

When analyzing GM composition, a lower abundance of Burkholderiaceae, Bacteroidales F082, and Parasutterella was observed in all HFD groups, confirming results from previous studies in mice fed a similar diet [27, 28]. Additionally, we observed that HFD also decreased the abundance of Coriobacteriales and Muribaculum and increased that of Lachnospiraceae, compared to C group, but these observations were not observed in the HFCAB group. These results confirmed those from previous studies [29].

In our study, Calafate administration prevented the HFD-induced increase in Tyzzerella 3, Lachnospiraceae NK4A136 group y Desulfovibrio genera. Previous studies have described that higher Tyzzerella abundances correlated with body and liver weight, perirenal fat, mesenteric fat, triglycerides, blood glucose, HOMA-IR, TNF-α, and IL-1β [30]. HFD has also been associated with an increase in bacteria from the Lachnospiraceae NK4A136 genus, this change being prevented by the 10 week-administration of a rosemary extract [31]. In contrast to our results, the relative abundance of bacteria from the Desulfovibrio genus has been shown to increase with the administration of polyphenols [26]. However, a diminution of Desulfovibrio might be considered as beneficial for the host’s health, since they are sulfate-reducing bacteria that are involved in production of H2S, a potentially harmful metabolite [32].

Our research shows that the administration of Calafate in the HFC group prevented the decrease in the relative abundance of the Prevotellaceae family and the genera Prevotellaceae UCG-003, Rikenellaceae RC9 gut group induced by HFD. Similarly, the administration of procyanidins (100 mg/kg body weight) for 12 weeks to HFD-fed mice has been shown to increase the abundance of Rikenellaceae RC9 [33]. Furthermore, Calafate extract administration increased the abundance of the Ruminococcaceae UCG-010 genus, compared to the HF group, therefore confirming previous studies in HFD-fed animals [34].

Despite the fact that the HFCAB group was treated with antibiotics, it presented a higher abundance of Akkermansiaceae and Peptococcaceae families, and A. muciniphila species, compared to the other groups. Previous investigations with animals treated with broad-spectrum antibiotics have reported similar expansion of A. muciniphila [35]. Increased abundance of Akkermansiaceae and Peptococcaceae has also been observed in germ-free animals after fecal transplant [36, 37], and it has been proposed that it could be due to an imbalance in the immune system of these animals [38].

Our study has some limitations. First, it is not possible to determine the mechanisms that best explain the results. Furthermore, the antibiotic combinations used in the HFCAB group cannot completely deplete GM, and we cannot reject the idea that these antibiotics may have an effect on thermogenesis. Finally, the animals were housed at room temperature throughout the experiment. Recent research has indicated that housing mice at room temperature (below their thermoneutrality) can lead to a significant increase in metabolic rate since mice must defend their core temperature by activating both shivering and non-shivering thermogenesis [39]. Despite these limitations, our results confirm the acute administration of Calafate extract induce the expression of thermogenic markers, prevents alterations in mitochondrial cristae and modulates GM in mice fed a high-fat diet.

This study was approved by the Institutional Committee for the Care and Use of Animals (CICUA) of the University of Chile, Santiago, Chile (protocol 17049-MED-UCH, accepted on June 27, 2017). All procedures were performed following the institutional policies and guidelines of the CICUA of the Universidad de Chile.

The authors have no conflicts of interest to declare.

The National Agency for Research and Development (ANID, Chile) funded this work (Grant FONDECYT#1171550 to D.F.G-D.).

Lissette Duarte: contributed to the design and the implementation of the research, the analysis of the results, and the writing of the manuscript; Vannesa Villanueva, Robert Barroux, Juan Francisco Orellana, Carlos and Poblete-Aro: contributed to the data curation of the research and the writing of the manuscript; Martin Gotteland: contributed to the analysis of the results and the writing of the manuscript; Mauricio Castro and Fabien Magne: contributed to the data curation, the analysis of the results, and the writing of the manuscript; Diego Garcia-Diaz: conceived the study and were in charge of direction and planning of project, directed the project and were involved in planning and supervision the work, and contributed to the analysis of the results and the writing of the manuscript.

All data generated in this study are included in this article. For more information, please contact the corresponding author.