The main ingredients of CDT include Fu brick tea (50%), green tea (25%), Ilex latifolia Thunb (kuding tea) (10%) and Momordica grosvenori (Luo Han Guo) (15%), which were purchased from Hunan Chu Ming Tea Industry Co., Ltd. (Hunan, China). The preparation of CDT aqueous extracts was as follows: low concentration CDT (6 mg/mL, CDT_L group), 1.8 g CDT was soaked in 300 mL boiling water for 25 min, then filtered out the tea residue and allowed the mixture to cool; high concentration CDT (12 mg/mL, CDT_H group), 3.6 g CDT was soaked in 300 mL boiling water for 25 min, then filtered out the tea residue and allowed the mixture to cool.

Mouse diet preparation was performed by Trophic Animal Feed High-tech Co., Ltd. (Jiangsu, China). Diet composition and calorie levels in mice are shown in Supplementary Tables S1, S2. In addition, 36 male C57BL/6J mice (6 weeks old) were purchased from Hunan Slake Jingda Experimental Animals Co., Ltd. (Hunan, China). All mice were maintained at a controlled temperature of 22 ± 2 °C, humidity of 65 ± 5%, and a 12-h light/dark cycle. All the mice had ad libitum access to food and water. After 1 week of adaptive feeding, 6-week-old male C57BL/6J mice were randomly divided into four groups: the Con (including the common diet group, n = 9), HFD (HFD group, n = 9), CDT_L (HFD + low-concentration CDT group, n = 9), and CDT_H (HDF + high-concentration CDT group, n = 9) groups. Mice in the CDT_L and CDT_H groups all received drinking CDT in sterile plastic bottles, which were replaced every 2 days with freshly made low- and high-concentration CDT, with the surplus being collected and measured. Food consumption and body weight were recorded once a week for 18 weeks. At the end of the experiment, the mice were anesthetized with tribromoethanol after 8 h of fasting. The cecum contents, liver, serum, perirenal, epididymal, subcutaneous, and brown adipose tissues were collected for follow-up experiments, immediately frozen in liquid nitrogen, and stored at −80 °C. The experimental procedures were performed following the Animal Care and Use Guidelines of China and were approved by the Animal Care Committee of Hunan Agricultural University and the Use Committee at HUNAU (No. 43322108).

After 18 weeks, an oral glucose tolerance test (OGTT) was performed as previously described (31). Briefly, the mice were fasted for 6 h, followed by oral administration of glucose (2.0 g/kg body weight). Then, the blood glucose concentrations were analyzed at 0, 30, 60, 90, and 120 min after the oral administration of glucose, and a glucose concentration-time plot was prepared to offer the integrated areas under each curve (AUC) for OGTT. The glucose levels were measured using an ACCU-CHEK glucose meter (Roche, Shanghai, China). The data were recorded and analyzed.

The serum levels of triglyceride (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), alkaline phosphatase (ALP), and aspartate aminotransferase (AST) were determined using a model BS-240 automatic biochemical analyzer (Shenzhen Mindray Bio-Medical Electronics, Shenzhen, Guangdong, China). Liver TG, TC, and FFA levels were assessed using commercial kits (Beijing Box Shenggong Technology Co., Ltd., Beijing, China). The serum levels of insulin, resistin, leptin, and glucagon-like peptide 1 (GLP-1) were assessed using commercial kits (Supplementary Table S3). Additionally, fasting blood glucose content was measured using an Accu-Chek Performa glucose meter (Roche, Shanghai, China). The fasting blood glucose and fasting insulin values were used to calculate the homeostasis model assessment of insulin resistance (HOMA-IR) index as follows: HOMA-IR = (fasting glucose × fasting insulin)/22.5 (32).

Subcutaneous adipose, brown adipose, and liver samples were cut into small pieces, dehydrated, cleared, embedded in paraffin wax, and sectioned after fixation in 10% neutral-buffered paraformaldehyde. Sections (5 mm) of each sample were prepared using hematoxylin and eosin (H&E) staining methods. To validate the vacuolization analysis and quantify lipid droplets, frozen liver samples were sectioned, stained with 0.2% oil red O in 60% isopropanol, and washed three times with phosphate-buffered saline (PBS). Images of stained liver tissues were collected using an orthographic light microscope (Nikon).

Brown adipose tissue was fixed in 4% paraformaldehyde and embedded in paraffin. Each sample was sliced into five sections discontinuously and then used for evaluation of relative brown adipose tissue by immunohistochemistry. After antigen repair and blockage of the endogenous peroxidase, sections were washed in PBS three times for 10 min and then incubated overnight at 4 °C with a rabbit antibody specific for GLP-1 (1:200, AiFang biological, AF11181). Tissue sections were washed three times in PBS for 10 min and incubated for 2 h at room temperature with the secondary antibody (HRP-Polymer anti-mouse/rabbit universal IHC kit, AiFang biological, AFIHC001, Changsha, China). After washing in PBS three times, tissue sections were stained with the chromogenic substrate 3,3′-diaminobenzidine (DAB), and the positive cell number was observed under the microscope. The number of positive cells was recorded at every high magnification, and the average of each sample was calculated. Sections from five animals from different treatment groups were analyzed (33).

Total RNA was isolated using TRIzol reagent (Accurate Biology, Hunan, China). RNA was then converted into complementary DNA (cDNA) using the Evo M-MLV Reverse Transcription Kit (Accurate Biology). Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) assays were performed using the SYBR Green Premix Pro Taq HS qPCR Kit (AG11701; Accurate Biology, Hunan, China) and on a high-throughput PCR instrument (WaferGen Biosystems, Fremont, CA, USA). Thermal cycling conditions were as follows: one cycle at 95 °C for 10 min, 95 °C for 30 s, and 60 °C for 30 s. Expression levels of genes were normalized to β-actin expression. The quantitative changes in gene expression were calculated using the 2–ΔΔCt method, where Ct = Ct (target gene) – Ct (β-actin). The primer sequences are listed in Supplementary Table S4.

The intestinal microbiota diversity detection experiment was commissioned by Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China) based on the Illumina MiSeq PE300 platform. The specific operation was the same as that in previous research methods (34). Bacterial genomic DNAs in feces (n = 7 per group) were sequenced by Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). Alpha diversity, including the Shannon and Sobs indexes, was calculated using Mothur (version 1.30.2). Comparative analysis of samples between groups was performed using partial least squares discriminant analysis (PLS-DA). Spearman's correlation analysis was conducted to explore the correlation between bacterial flora and physiological factors.

The results were presented as the mean ± standard error of the mean (SEM). Statistical Package for the Social Sciences (SPSS) (version 25.0; IBM SPSS, Chicago, IL, USA) and GraphPad Prism (version 9.0; GraphPad Software, San Diego, CA, USA) were used for the statistical analyses. Data were tested for normality using the Shapiro–Wilk test. Normally distributed data were analyzed using a one-way analysis of variance (ANOVA). Non-normally distributed data were analyzed using the Kruskal–Wallis rank-sum test. Statistical significance was set at P < 0.05, while P < 0.01 was considered highly significant.

C57BL/6 mice were supplemented with CDT under continuous HFD feeding for 18 weeks to investigate the anti-obesity effects of CDT in HFD-fed mice. As shown in Figures 1A, B, the CDT_L and CDT_H groups had significantly lower body weights than that of the HFD group (P < 0.01), which decreased by 15 and 16%, respectively. Furthermore, as expected, the body fat rate decreased by 44% in the CDT_L group, while the body fat rate in the CDT_H group decreased by 38% (Figure 1C). More notably, the difference in body weight and body fat rate was not ascribed to reduced food intake and water consumption, since no significant difference was found in food intake and water consumption between all groups (Figures 1D, E). As shown in Figure 1F, after drinking low- and high-concentrations of CDT, the fasting blood glucose levels of the CDT_L and CDT_H groups were reduced to varying degrees (P < 0.05 and P < 0.01, respectively). Similar trends were observed in the results of the OGTT among the groups; the OGTT was significantly reduced in both the CDT_L and CDT_H groups (Figure 1G, P < 0.01). Moreover, HFD treatment caused notable increases in the mice's insulin, resistin, HOMA-IR, and leptin values, which were restored in the CDT_L and CDT_H groups (Figures 1H–K). Compared with the Con group, drinking low- and high-concentrations of CDT significantly increased the content of GLP-1 (Figure 1L, P < 0.05). By measuring serum lipid markers, the serum levels of TG, TC, HDL-C, and LDL-C were higher in HFD-fed mice than in the CDT_L and CDT_H groups (Figures 1M–P).

When comparing obesity-related traits, we demonstrated that representative histological slices of subcutaneous adipose tissue, as presented in Figure 2A, suggested that CDT significantly lowered the mean adipocyte size in HFD-fed mice. Interestingly, the CDT_L group had a better effect on reducing the size of adipocytes than the CDT_H group. In addition, CDT was able to reduce the size of brown adipose cells on the HFD, as shown in Figure 3A, and the effect of the CDT_H group was better than that of the CDT_L group. Previous studies have reported that the body upregulates the expression of the brown adipose tissue-specific gene, uncoupling protein 1 (Ucp-1), enhances energy expenditure, and promotes browning of subcutaneous adipose tissue, preventing weight gain (35). Therefore, to determine whether the increased energy expenditure in the CDT_L and CDT_H groups was associated with adaptive thermogenesis, we assessed the effects of CDT administration on brown adipose tissue. Immunohistochemical analysis of Ucp-1 revealed that CDT administration promoted thermogenesis of brown adipose tissue compared to that in the HFD group (Figures 3B, C, P < 0.05). To further understand the effects of CDT intervention on lipid metabolism, we analyzed the expression of relevant genes in the white and brown adipose tissues. We observed lower mRNA expression of thermoregulatory markers and lipid metabolism-related genes, including Ucp-1, peroxisome proliferation-activated receptor-gamma coactivator 1 alpha (Pgc-1α), peroxisome proliferator-activated receptor gamma (Ppar-γ), PR domain containing 16 (Prdm-16), cell death-inducing DNA fragmentation factor alpha-like effector A (Cidea), and fibroblast growth factor 21 (Fgf-21), in the white adipose tissue of the HFD group than in the Con group. However, the CDT_L and CDT_H groups expressed higher mRNA levels of these thermoregulatory markers and lipid metabolism-related genes than the HFD group (Figures 2B–G, P < 0.01). Furthermore, to examine the changes in mRNA expression of thermoregulatory markers and lipid metabolism-related genes in brown adipose tissue by HFD, the mRNA expression of thermoregulatory markers and lipid metabolism-related genes was determined (Figures 3D–K). The expression of Ucp-1, Ppar-γ, Pgc-1α, Prdm-16, CREB-regulated transcription coactivator 2 (Crct2), carnitine palmitoyltransferase 1 (Ctp1), nuclear respiratory factor 1 (Nrf1), and Nrf2 was lower in HFD-fed mice than in the Con group. However, the CDT_L group exhibited a significantly upregulated mRNA expression of Ucp-1, Ppar-γ, Pgc-1α, and Ctp1 (P < 0.01) and upregulated mRNA expression of Nrf1 and Nrf2 compared with the HFD group (P < 0.05). The CDT_H group exhibited a significantly upregulated mRNA expression of Ucp-1, Pgc-1α, and Prdm-16 (P < 0.01) and upregulated mRNA expression of Ppar-γ, Ctp1, and Nrf1 compared with the HFD group (P < 0.05).

As shown in Figure 4A, H&E staining of liver sections revealed that HFD-induced indistinct cell boundaries, denaturation, irregular damaged cells with cytoplasmic vacuolation and that the CDT_L and CDT_H interventions improved this outcome. Oil red O staining revealed that CDT_L and CDT_H alleviated HFD-induced hepatic steatosis, and the effect of CDT_L was greater than that of CDT_H (Figures 4B, C). Liver weight was significantly increased in HFD-fed mice compared to that in the Con group, whereas CDT_L and CDT_H intervention remarkably lowered liver weight in mice on HFD (Figure 4D, P < 0.05 and P < 0.01, respectively). CDT_L and CDT_H also markedly reduced the levels of hepatic TC and TG in HFD-fed mice (Figures 4E, F; P < 0.01 and P < 0.05, respectively). Furthermore, CDT_L significantly reduced liver FFA levels in HFD-fed mice (Figure 4G, P < 0.05). The elevated levels of ALT in the HFD group were significantly ameliorated by CDT_H treatment (Figure 4I). In addition, Figures 4H, J demonstrates that CDT_L and CDT_H enhanced the increase AST and ALP levels in the HFD group, indicating that CDT_L and CDT_H exert protective effects on hepatotoxicity. Furthermore, to examine the changes in hepatic lipid metabolism regulated by HFD, the expression of lipid metabolism-related genes was determined (Figure 4K). The expression of genes, including adipose triglyceride lipase (Atgl), was lower in the HFD-fed mice than in the Con group. However, CDT_L and CDT_H significantly upregulated the mRNA expression of Atgl compared to that in the HFD group (P < 0.01). At the same time, the expression of sterol regulatory element-binding protein 1c (Srebp-1c), Fas, and acetyl-CoA carboxylase 2 (Acc2), which are related to lipid synthesis in the liver, was upregulated under HFD conditions. However, CDT_L and CDT_H significantly reduced the mRNA expression of Srebp-1c, Fas, and Acc2 compared with the HFD group. In addition, the expression of Acc1 is significantly reduced in the CDT_L group compared to that in the HFD group (P < 0.01).

The effects of HFD and CDT on the gut microbiota of mice were analyzed in the current study. After removing the unqualified sequence, as shown in Figure 5A, the Con group displayed 531 operational taxonomic units (OTUs), whereas the HFD, CDT_L, and CDT_H groups displayed 534, 558, and 578 OTUs, respectively. As shown in Figure 5B, the Shannon index at the OTU level was not significantly different among the four groups, but there was an upward trend in the CDT_L group compared with the other three groups. The Sobs index of OTU level results demonstrated a significant increase in the CDT_H group compared with the HFD group (P < 0.05). In addition, it was observed that there was a tendency to reduce the Sob index of the HFD group compared with the Con and CDT_L groups (Figure 5C). Moreover, PLS-DA analysis revealed differences between the HFD, CDT_L, and CDT_H groups (Figure 5D).

To explore differences in gut microbial composition, the relative abundance of bacterial communities at the phylum and genus levels was analyzed. At the phylum level, the gut flora of the four groups was dominated by Firmicutes, Bacteroidota, Desulfobacterota, and Actinobacteria (Figure 6A). As shown in Figure 6B, at the phylum level, Desulfobacterota accounted for 12.25% of the intestinal flora in the Con group, and this type of bacteria constituted 5.64% of the microbial profile in the CDT_L group. Actinobacteria accounted for 8.95% of the intestinal flora in the CDT_H group and 3.17% of the microbial profile in the CDT_L group (Figure 6C). At the order level, an increased abundance of Clostridiales was observed in the CDT_L group (Figure 6D). In addition, high dietary fiber intake significantly increased the relative abundance of Rhizobiales in HFD-fed mice (Figure 6E). High concentrations of CDT significantly increased the relative abundance of Clostridia_vadinBB60_group, compared with those in both the HFD and CDT_L groups (Figure 6F). Compared with the HFD group, the relative abundance of Burkholderiales in the CDT_H group decreased (P < 0.05) (Figure 6G). At the genus level, the Con group possessed significantly higher levels of Desulfovibrio than did the CDT_L group (Figure 6H). Meanwhile, high concentrations of CDT led to an increase in the relative abundance of Bifidobacterium compared to the CDT_L group (Figure 6I). In addition, the low concentration of CDT markedly increased the relative abundances of multiple genera, including norank_f__Eubacterium_coprostanoligenes and Ruminococcus, when compared to the HFD group (Figures 6J, K). The CDT_H group possessed significantly higher levels of Lactococcus but lower levels of Parasutterella than those in the HFD group (Figures 6L, M).

Spearman's correlation analysis was performed to understand the association between differentially enriched microbes and obesity-related traits (Figure 7). Nineteen and 12 OTUs were negatively and positively correlated with obesity-related traits, respectively. Bifidobacterium (OUT 91) was positively correlated with ALP. Low-concentration CDT significantly increased the relative abundance of Bifidobacterium compared to that in the CDT_H group (Figure 6I). norank_f__Eubacterium_coprostanoligenes (OUT 587) was observed to be negatively correlated with the weight gain, ALP, and AUC of the OGTT. Meanwhile, low concentrations of CDT led to an increase in the relative abundance of norank_f__Eubacterium_coprostanoligenes compared to the Con and HFD groups (Figure 6J). The gut flora of the two groups were dominated by norank_f__Muribaculaceae and Lachnospiraceae_NK4A136_group at the genus level. The norank_f__Muribaculaceae was negatively correlated with body weight, fat percentage, weight gain, AUC of OGTT, liver TC, HDL-C, LDL-C, HOMA-IR, insulin, resistin, and leptin, including OUTs 591, 499, and 593, OTU 654, and OUTs 448, 16, 483, 33, 319, 458, and 480. Similarly, the Lachnospiraceae_NK4A136_group was negatively correlated with body weight, fat percentage, weight gain, AUC of OGTT, liver TC, HDL-C, LDL-C, AST, ALP, HOMA-IR, blood glucose, insulin, resistin, and leptin, including OUTs 56, 102, 336, 42, and 581. Taken together, these results suggest that the low-concentration CDT intervention modulates HFD-induced gut microbiota dysbiosis in mice.