3.1 CBPs prevents obesity, liver steatosis and improves dyslipidemia in HFD-fed rats
After two months of continuous feeding high fat diet, the body weight of rats in high fat diet group and control group were 594 ± 47.73 g and 476.72 ± 32.95 g, respectively. There was significant difference between the two groups, which indicated that the obesity model of rats was successfully established. The body weight of AM group rats decreased continuously during CBPs treatment. Weight gain of AM group rats has a significant difference compared with HF group rats (P<0.001) (Fig. 1A, B). Notedly, the body weight of HF and SV group rats increased slowly, and weight gains were 5.29% and 2.72%, respectively, with significant difference (P<0.05) (Fig. 1B). Compared with the HF group rats, the weight of visceral adipose tissues (eWAT, pWAT and mWAT), subcutaneous adipose tissue (iWAT) and liver reduced in AM group rats and SV group rats (except eWAT) (Fig. 1C, D). There were no significant differences in weight of iBAT, heart, kidney, spleen, lung, pancreas and testicle among the three groups (Fig. 1D and Supplementary Fig.1). Overall, CBPs treatment tended to prevent weight gain by reducing the weight of liver, visceral and subcutaneous adipose tissues in HFD-fed rats. Simvastatin treatment also inhibited weight gain slightly in HFD-fed rats. However, its effect of improving obesity was inferior to CBPs treatment. There were significant difference in liver TC and TG concentration of AM and HF group rats (p<0.05) (Fig. 1G, K). Similarly, H&E staining of liver and adipose tissues also showed obese rats treated with CBPs significantly reduced the hepatic fat droplets and adipocyte size compare with HF group rats (Fig. 1E). Simvastatin treatment could decreased liver TC concentration, whereas it had no effect on reducing liver TG concentration in HFD-fed rats (Fig. 1G, K). Besides, simvastatin treatment also improved the fat accumulation and reduced adipocyte size in liver and adipose tissues, which was less effective than CBPs treatment (Fig. 1E).
The serum TC, TG and LDL-C increased significantly and the serum HDL-C decreased in HF group rats compared with the control group rats (P<0.001) within 40 days of HFD feeding (Fig. 1F-I), suggesting that rats fed with high-fat diet could be induced to develop hyperlipidemia. During the CBPs treatment, serum TC, TG and LDL-C decreased gradually and there were significant differences compared with HF group after 20, 30 and 20 days, respectively (Fig. 1F-H). Serum HDL-C increased gradually within 40 days of CBPs treatment, which was a significant difference between AM group and HF group after 30 days (Fig. 1I). The results manifested that CBPs treatment could improve hyperlipemia by reducing serum TC, TG, LDL-C and increasing HDL-C concentrations in HFD-fed rats. Simvastatin treatment could also significantly improve hyperlipidemia, which was more effective to reduce serum TC and LDL-C than CBPs treatment. There were significant differences in serum TC and LDL-C between SV and HF group rats after 10 days (Fig. 1F, H). However, the effect of CBPs treatment on reducing serum TG and increasing HDL-C were better than that of simvastatin treatment in HFD-fed rats (Fig. 1G, I).
3.2 CBPs alters gut microbial composition in HFD-fed rats
We analyzed the fecal microbial composition of HF, AM, SV group rats after 10, 20, 30 and 40 days. ACE, Chao1, shannon and simpson indexes were examined for the richness and alpha-diversity of the gut microbiota. The HF group rats revealed significantly higher ACE and Chao1 indexes after 20 days. There was no significant differences of shannon and simpson diversity indexes in HF group within 40 days (Table 2). Meanwhile, the ACE, Chao1, shannon and simpson indexes were no significant differences after treatment with CBPs and simvastatin (Table 2).
CBPs supplementation had a greater effect on gut microbial composition. At the phylum level, the relative abundance of Bacteroidetes reduced and the relative abundance of Firmicutes, F/B ratio increased in HF group rats, which were no significant change after 20 days (Fig.2A). Conversely, the relative abundance of Bacteroidetes and Verrucomicrobia were gradually increased in the AM group rats within 40 days, while the relative abundance of Firmicutes and F/B ratio were suppressed markedly (Fig.2A). The relative abundance of Proteobacteria decreased significantly within 20 days in HF group rats, which was no significant variation after 20 days. In SV group rats, except for Proteobacteria, there were no significant change in relative abundance of Bacteroidetes, Firmicutes, Verrucomicrobia and F/B ratio within 40 days (Fig.2A).
At the genus level, the gut microbial composition showed similar trends to the phylum level. The relative abundance of Firmicutes phylum (Lachnospiraceae_NK4A136_group, Lachnoclostridium), Proteobacteria phylum (Desulfovibrio) were decreased and the relative abundance of Bacteroidetes phylum (Bacteroides, Prevotella), Firmicutes phylum (Romboutsia), Verrucomicrobia phylum (Akkermansia) were increased gradually in AM group rats within 40 days (Supplementary Fig.2). However, the relative abundance of Firmicutes phylum (Lachnospiraceae_NK4A136_group, Clostridium) were higher and Bacteroidetes phylum (Bacteroides, Prevotella), Verrucomicrobia phylum (Akkermansia) were lower in HF group rats compared with AM group rats after 40 days (Supplementary Fig.4). Except for the increasing in the relative abundance of genus Clostridium, there was no significant change in other genus within 40 days of simvastatin treatment (Supplementary Fig.2). Furthermore, the linear discriminant analysis (LDA) effect size (LEfSe) was used to identify the biomarkers with significant differences between the two groups. After 40 days, compared with the AM group rats, the relative abundance of Firmicutes phylum (Romboutsia, Ruminococcaceae, Turicibacter, UBA1819, Anaerotruncus) and Actinobacteria phylum (DNF00809) were altered significantly in HF group rats. However, the the relative abundance of Bacteroidetes phylum (Bacteroidia, Bacteroidales, Muribaculaceae, Bacteroidaceae, Bacteroides, Prevotella) and Proteobacteria phylum (Alphaproteobacteria, Rhodospirillales) had significant differences in AM group rats compared to HF group rats (Fig.2C). Simultaneously, LEfSe analysis elucidated the genus level differences such that HF group rats was more abundant in species of Christensenellaceae compared with SV group rats, whereas there was only one genus (Paenalcaligenes) had significant differences in SV group rats compared with HF group rats (Fig. 2D).
3.3 CBPs changes serum BAs pool, which is related in gut microbial composition
BAs synthesis is an important pathway for catabolism of cholesterol and is closely regulated by complex mechanisms that are not completely understood. BAs were considered as mediators of metabolism, alteration the BAs homeostasis will cause many diseases such as obesity, diabetes, nonalcoholic fatty liver disease and hyperlipemia [26]. We anticipated that CBPs treatment could shift the BAs pool in HFD-fed rats. As can be seen from Fig. 3A, the total serum BAs content of AM group rats increased 20 days ago and then decreased gradually after 20 days. Nevertheless, within 40 days of high-fat diet feeding, the total serum BAs content increased continuously in HF group rats. These results suggested that CBPs can significantly improve the shift of BAs pool which induced by HFD in obese rats. The total serum BAs content of SV group rats increased continuously and decreased slightly after 30 days. Furthermore, the relative content of cholic acid (CA), deoxycholic acid (DCA) and taurohyodeoxycholic acid (THDCA) were decreased gradually in AM group rats, while the relative content of chenodeoxycholic acid (CDCA), hyodeoxycholic acid (HDCA), ursodeoxycholic acid (UDCA) and β-muricholic acid (β-MCA) were enhanced in AM group rats (Supplementary Fig.3). Compared with AM group rats, the relative content of CA and DCA was higher and relative content of β-MCA and HDCA were lower in HF group rats after 40 days. In addition, the relative content of UDCA increased slightly and the relative content of TUDCA and THDCA decreased in SV group rats within 40 days.
Correlation coefficients between the relative content of serum BAs and the relative abundance of gut bacteria at genus-level were shown in Table 4. Several BAs correlated with specific bacterial genera. Bacteroides was positively correlated with CDCA, HDCA and negatively correlated with DCA, GCA. Similarly, Prevotella was positively correlated with CDCA, HDCA, β-MCA and negatively correlated with DCA, GCA, TUDCA. Interestingly, Acetitomaculum and Prevotella have the same trend of association with BAs. In addition, β-MCA positively correlated with Akkermansia. On the contrary, Desulfovibrio was negatively correlated with β-MCA and CDCA.
3.4 CBPs regulates gene expression in liver and adipose tissues of HFD-fed rats
To further explore the molecular mechanism of CBPs improving obesity in HFD-fed rats, we evaluated the gene expression of lipogenesis, lipolysis and BAs metabolism in liver and adipose tissues. In the liver tissue, compared with HF group rats, AM group rats significantly enhanced the mRNA expression of peroxisome proliferator-activated receptor α (PPARα), peroxisome proliferator-activated receptor γ (PPARγ), small heterodimer partner (SHP), G protein-coupled bile acid receptor (TGR5), fibroblast growth factor 15 (FGF15), fibroblast growth factor 4 (Fgfr4), bile salt export protein (BSEP) and downregulated the mRNA expression of cholesterol-7a-hydroxylase (CYP7A1) (Fig.3C). The results indicated that CBPs treatment could alleviate the disorder of hepatic BAs metabolism and fat accumulation. Simvastatin treatment also partially improved hepatic BAs metabolism by up-regulating SHP and TGR5 gene expression (Fig.3C).
In the eWAT, CBPs treatment markedly downregulated the mRNA expression of PPARα, PPARγ, acetyl-coenzyme A carboxylase 1 (ACC1), sterol regulatory element binding protein-1c (SREBP-1c), CCAAT enhancer binding protein α (C/EBPα), fatty acid synthetase (FAS) and enhanced the mRNA expression of hormone-sensitive lipase (HSL) in AM group rats compared with HF group rats (Fig.3D). In accordance with the eWAT, PPARγ, ACC1, C/EBPα, FAS were dramatically downregulated and HSL was upregulated in the iWAT after CBPs treatment (Fig.3E). Simvastatin treatment reduced mRNA abundance of ACC1 and increased mRNA abundance of PPARα in the eWAT compared with HF group rats (Fig.3D). Meanwhile, the mRNA expression of PPARα, PPARγ, ACC1, C/EBPα and FAS were downregulated slightly to improve fat accumulation in the iWAT after simvastatin treatment (Fig.3E). In the iBAT, compared with HF group rats, CBPs treatment positively regulated the mRNA expression of peroxisome proliferator-activated receptor γ co-activator 1α (PGC-1α), PPARγ and upregulation of uncoupling protein 1 (UCP1) in AM group rats, while simvastatin treatment had no similar effect in SV group rats (Fig.3F). Consequently, we could conclude that CBPs can improve lipid metabolic syndrome in HFD-fed rats by regulating the related mRNA expression of lipogenesis and lipolysis in the WAT and modulate energy homeostasis and thermogenesis in the iBAT.
3.5 Fecal microbiota transplantation (FMT) from CBPs-treated rats remodels gut microbiota and improves dyslipidemia in HFD-fed rats
We investigated the FMT from CBPs-treated rats remodeled gut microbiota and improved lipid metabolism in HFD-fed rats. As shown in Fig.4 A-B, there were no significant difference in body weight and weight gain between the FMT-HF and FMT-AM group rats within 30 days. And the weight of liver, kidney, spleen, iWAT, eWAT, pWAT and iBAT were no significant difference between the FMT-HF and FMT-AM group rats (Supplementary Fig.5). However, FMT from CBPs-treated rats could significantly reduce serum TC, TG, LDL-C and increase HDL-C in FMT-AM group rats compared with FMT-HF group rats (Fig.4 E-H). In liver, the concentration of TC and TG showed no significant difference in the two groups rats (Fig.4 C-D).
Furthermore, to reveal the effects of FMT on the gut microbial structure, we sequenced the fecal bacterial 16S rRNA after 10, 20 and 30 days in FMT-HF group rats and FMT-AM group rats. The ACE and Chao1 indexs were increased gradually, while the shannon and simpson indexes did not change significantly within 30 days in FMT-AM and FMT-HF group rats (Table 3). At the phylum level, FMT from CBPs-treated rats tended to increase the relative abundance of Bacteroidetes, Verrucomicrobia and Epsilonbacteraeota but decrease the relative abundance of Firmicutes and Actinobacteria within 30 days. Conversely, the relative abundance of Firmicutes, Actinobacteria were higher and Bacteroidetes, Verrucomicrobia and Epsilonbacteraeota were lower in FMT-HF group rats compared with FMT-AM group rats (Fig. 5A). The F/B ratio was increased dramatically in FMT-HF group rats, whereas FMT from CBPs treatment rats reversed this trend significantly after 30 days.
At the genus level, the relative abundance of Bacteroides, Prevotella and Akkermansia was higher, while the relative levels of Blautia and Streptococcus were markedly lower in FMT-AM group rats compared with FMT-HF group rats (Supplementary Fig.6). In FMT-HF group rats, the relative abundance of Prevotella, Phascolarctobacterium was reduced and the relative abundance of Lactobacillus, Eubacterium was increased gradually. The LEfSe analysis results indicated the relative abundance of Firmicutes phylum (Bacilli, Lactobacillales, Lactobacillaceae, Lactobacillus, Erysipelotrichia, Erysipelotrichales, Erysipelotrichaceae, Allobaculum, Blautia, Eubacterium, Ruminococcus, Clostridium) in FMT-HF group rats was significantly increased compared with FMT-AM group rats (Fig. 5C). The Bacteroidetes phylum (Bacteroidia, Bacteroidales, Bacteroidaceae, Prevotellaceae, Prevotella, Muribaculaceae), Verrucomicrobia phylum (Verrucomicrobiae, Verrucomicrobiales, Akkermansiaceae, Akkermansia) and Firmicutes phylum (Negativicutes, Selenomonadales, Acidaminococcaceae, Phascolarctobacterium) were identified by LEfSe as discriminative taxa in FMT-AM group rats compared with FMT-HF group rats (Fig. 5C).