3 Results
3.1 Effects on body weight and fat
Fig.1a-d shows the body weight (the initial weight, weight of modeling, and final weight) and body fat percentage in rats. There was no significant difference in the initial body weight between groups (Fig.1a) (P > 0.05). During the treatment period, the body weights of rats in OC, DG, VD12.5, and DG10VD12.5 groups were significantly increased compared with NO group (Fig.1b) (P < 0.05). However, the increment in body weight was lower in VD and DG group (278.22, 280.38, P< 0.05), and further lower in DG10VD12.5 group (260.28, P < 0.05) compared with the OC group (298.33) (Fig.1c). For the body fat percentage, rats under DG10D12.5 treatment had the lowest value (4.34) and had lower values under VD and DG treatment (6.03, 5.78) compared with the OC group (9.04, P < 0.05 for all).(Fig.1d)
3.2 The effects on serum lipid levels and metabolism enzyme activity
The effects of DAG and VD on serum lipids are shown in Fig.2a-d. The obesity groups (OC, DG, VD12.5 and DG10VD12.5 treatments) had significant increases in the serum levels of TC, TG, and LDL, whereas reduction in HDL compared with those in NO group (P < 0.05). The DG group or VD group and DG plus VD group had significant influences on the serum lipids compared with those in OC group (P< 0.05). For HDL, the highest value was found in rats under the DG10VD12.5 treatment (3.10), and the VD12.5 and DG10 treatment had relatively lower value (2.84, 2.73) compared with the OC group (2.02) (Fig.2c). For TG, TC, and LDL, the rats under the DG10VD12.5 treatment had the lowest values (1.07, 3.25, and 3.16) compared with the OC group (1.81, 4.53, and 4.07) (P < 0.05 for all) (Fig.2a, 2b and 2d).
The effects of DAG and VD on the metabolism enzyme activity in the liver are shown in Fig.3a and 3b. The obesity (OC, DG, VD12.5, and DG10VD12.5 treatments) resulted in significant decreases in the AST and ALT activity compared with those in the NO group (P < 0.05). There were significant differences on the AST and ALT activity between OC and DG10VD12.5 treatments (P< 0.05), however, there were no significant differences between OC and DG10 or VD12.5 group (P > 0.05). For AST, the highest value was found in rats under the DG10VD12.5 treatment (139.14), and the VD12.5 and DG10 treatment had relatively lower value (133.57, 133.81) compared with the OC group (129.42) (Fig.3a). For ALT, the highest value was found in rats under the DG10VD12.5 treatment (49.77) ,and the VD12.5 and DG10 treatment had relatively lower value (45.79, 46.23) compared with the OC group (43.58) (P < 0.05 for all) (Fig.3b).
3.3 The effects on FAS/CPT1 mRNA expression
The effects of DAG and VD on the expression of genes involved in the PPARγ/NF-κBp65 pathway are shown in Fig.4a, Fig.4b. Compared with those in the NO group, the obesity significantly lowered the CPT1 value (P < 0.05), but increased the FAS value (P < 0.05). There were no significant differences on FAS and CPT1 between NO and DG10VD12.5 treatments (P < 0.05). For CPT1 mRNA, the highest value was found in the rats under the DG10VD12.5 treatment (0.90) compared with the OC group (0.46) (Fig.4a). For FAS mRNA, the rats under the DG10VD1.5 treatment had the lower value (1.15) compared with the OC group (1.61) (Fig.4b).
3.4 The correlation between FAS/CPT1 mRNA expression, lipid metabolism, and metabolism enzyme activity
The correlation coefficients between the FAS/CPT1 mRNA expression and the metabolism enzyme activity in the liver, and lipids in serum are listed in Table 2. The FAS mRNA expression was highly and positively correlated with the TC, TG, and LDL concentrations in serum (P < 0.01), but negatively correlated with the ALT and AST activity (P < 0.01). The CPT1 mRNA was highly and negatively correlated with the TC, LDL (P < 0.01), and TG concentrations (P<0.05), and positively correlated with the ALT and AST activity (P < 0.01).
3.5 Effects on microbiota in the caecum
The diversity indices of microbiota, such as operational taxonomic units (OUTs), ACE index, Chao 1 index, Simpson index and Shannon index, in the caecal content are shown in Fig. 5a - 5e. The OC group had significant lower values of the diversity indices compared with those for the other four groups (P < 0.05). For OTUs, the high-fat diet reduced OTUs compared with the NO group, and the DG10VD12.5 treatment (459.33) had highest values compared with the OC group (233.00) (Fig.5a). For ACE index, the high-fat diet reduced ACE index compared with the NO group, and the DG10VD12.5 treatment (421.46), the DG10 treatment (393.81) and the VD12.5 treatment (372.71) had higher values compared with the OC group (263.73) (Fig.5b). For Chao 1 index, the high-fat diet reduced Chao1 index compared with the NO group, and the DG10VD12.5 treatment (423.38) had highest values compared with the OC group (264.79) (Fig.5c). No significant differences in Simpson (Fig.5d) and Shannon (Fig.5e) indices for the microbiota were found between the NO group and the DG10VD12.5 treatment (P > 0.05).
The difference of the microbiota composition was explored using the NMDS analysis. Fig. 6 is the result of the NMDS analysis. The Stress value lesser than 0.2 indicates that the NMDS analysis has certain reliability. The distance between the OC group and the NO group was the largest, indicating a significant difference in microbial composition. The distance between DG10VD12.5 treatment and the OC group was closer to that between VD12.5 or DG10 treatment and the OC group.
The microbiota taxonomic distribution at the genus level is shown in Fig. 7a - Fig. 7f. Obesity significantly reduced the abundances of Bacteroides andLactobacillus, whereas increased Allobaculum ,Desulfovibrio , Lachnospiraceae , and Alloprevotellagenera compared with those for the NO group (P < 0.05). In general, the additions of VD or DAG or VD plus DAG increased the abundances of Bacteroides and Lactobacillus , while reduced the abundances ofLachnospiraceae ,Alloprevotella , Desulfovibrio , and Allobaculum . ForLactobacillus , the highest value was found in rats under the VD12.5 treatment (0.159) ,and the DG10VD12.5 treatment had relatively lower value (0.107) compared with the OC group (0.023) (P < 0.05 for all) (Fig.7d). For the other five bacterial genera, the best effects were found in rats under the DG10VD12.5 treatment (0.051, 0.066, 0.014, 0.034, and 0.01) compared with the OC group (0.110, 0.013, 0.032, 0.051, and 0.066) (Fig.7a, 7b, 7c, 7e and 7f).
The changes in the microbiota composition were also explored using the Line Discriminant Analysis Effect Size (LEfSe) analysis. Fig. 8a and Fig. 8b are the results of the LDA analysis and evolutionary branch diagram. With the LDA threshold set at 4, a total of twenty groups of biomarker with statistical differences were detected. The bacterial taxa with significantly higher abundance in the DG10VD12.5 group wereAcidaminococcaeae and Phascolarctobacterium , while for the DG10 treatment were Rikenellaceae- RC9-gut, Collinsella , and Faecalibacterium . In contrast, Campylobacterales ,Helicobacteraceae , Helicobacter , andEpsilonproteobacteria were higher in the OC group, whereasPrevotellaceae -UCG-003, Prevotella -2, Treponema -2,Runminococcaceae -UCG-005, Selenomonadales ,Veillonellaceae , Negativicutes, Spirochaetales andRoseburia were higher in the NO group.