Nuciferine was bought from MedChemExpress Co., Ltd. (Shanghai, China, purity > 98%). Standard EGCG was obtained from Beijing Solarbio Technology Co., Ltd. (Beijing, China, purity > 98%). Chitosan of medium molecular weight was purchased from Sigma–Aldrich (St. Louis, MO, USA, purity > 75%). Lipo3000 transfection reagent was acquired from Thermo Fisher Co., Ltd. TRIzol^® Plus RNA Purification Kit and Transcriptor First Strand cDNA Synthesis Kit were bought from Thermo Fisher (Waltham, MA, USA). Dual-luciferase reporter assay system was purchased from Promega Co., Ltd. (Madison, WI, USA).

The following instruments were used in this study: Zetasizer Nano ZS90 (Malvern Instruments Co., Ltd., Malvern, UK), MIKRO 22R centrifuge (Kirchlengern, NRW, Germany), OLYMPUS BX60 fluorescent microscope camera (OLYMPUS Corporation, Tokyo, Japan), transmission electron microscope (TEM) (Hitachi, Tokyo, Japan), and CFX384 multiplex real-time fluorescence quantitative PCR instrument (Bio-Rad, Hercules, CA, USA).

Nuciferine-loaded liposomes (NF-liposomes) were manufactured by solvent evaporation and ultrasonic homogenization following a previous method (17) with some modifications. In brief, 1 g of soybean phospholipids (Shanghai Taiwei Pharmaceutical Co., Ltd., Shanghai, China) and 0.2 g of cholesterol (Shanghai Titan Chemical Co., Ltd., Shanghai, China) were dissolved in trichloromethane and stirred with magnetic force. After complete distribution, 0.2 g of NF dissolved in methanol was added and the mixture was stirred evenly. The organic solvent was evaporated by rotation (30 rpm) at 30°C to form a light-yellow film, which was then dispersed with 0.05 M phosphate buffered saline (PBS). Ultrasonic crushing was performed in an ice bath for 6 min to obtain a translucent liposome solution. Finally, particle size, zeta potential, and polydispersity index (PDI) were analyzed by Zetasizer Nano ZS90. Encapsulation rate was determined by high-performance liquid chromatography (HPLC).

Chitosan–procyanidin microgels were prepared following the methods of Zhang et al. (18) and Zou et al. (19) with some modifications. Chitosan was dissolved in 1% acetic acid solution to a concentration of 6 mg/ml. The liposome solution loaded with NF was mixed with chitosan solution, and magnetic stirring was performed for 1 h. MLPE with purity of 70% and average polymerization degree of 6.7 was extracted in our lab. The two concentrations of MLPE (2 and 6 mg/ml) were added to the above solution (EGCG was dissolved in advance so that the final concentration was 2 mg/ml), and magnetic stirring was performed for at least 3 h to obtain NF-EGCG double-encapsulated microgel (NFEG-microgel). After overnight refrigeration, the encapsulation rate and stability of the two concentrations of MLPE microgels were compared. The results showed that the encapsulation rate was high and stable upon the addition of 2 mg/ml MLPE in the microgel. Thus, subsequent microgels were prepared with this concentration.

Thirty healthy male Wistar rats (180 ± 20 g) with special pathogen free were purchased from SLAC Laboratory Animal Co., Ltd. (Shanghai, China) with animal license number SCXK (Shanghai, China) 2017-0005. The rats were raised in standard conditions (temperature 25 ± 1°C, relative humidity 40–60%, and 12 h light/dark cycles) and had free access to food and water. Their dietary consumption was recoded daily. All animal procedures were permitted by the Laboratory Animal Ethics Committee (2022-005) of China Jiliang University (Hangzhou, China). After a week of acclimatization, the rats were randomly divided into the following six groups with five rats each: blank control (NC), HFD, NF-loaded microgel treatment with high-fat diet (HFD + NF), EGCG-loaded microgel treatment with HFD (HFD + EGCG), NF and EGCG co-loaded microgel treatment with HFD (HFD + NFEG), and empty microgel treatment with HFD (HFD + MG). The NC group received commercial normal diet, and the remaining groups received HFD purchased from Fanbo Animal Feed Biotechnology Co., Ltd. (Shanghai, China) and consisted of 10% lard, 10% yolk powder, 6% casein, 3% maltose, 2% cholesterol, 1.2% premix, and 0.1% cholate. The rats in HFD + NF group intragastrically received NF at 20 mg/kg/day, those in HFD + EGCG group intragastrically received EGCG at 20 mg/kg/day (22), and those in HFD + NFEG group intragastrically received the same NF (20 mg/kg/day) and EGCG (20 mg/kg/day) for 8 weeks (23), The volume of the treatments was 10 ml/kg of rat body weight. The rats in NC and HFD + MG groups intragastrically received the same volume of water and empty microgel, respectively, for 8 weeks. Food intake was recorded daily, and body weight was monitored weekly. Blood samples were collected from the tail every 2 weeks and then centrifugated at 1,500 × g for 15 min at 4°C to separate the serum. The biochemical indexes of the serum were detected using the corresponding kits. Triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL-C), and low-density lipoprotein (LDL-C) kits were purchased from Thermo Fisher Co., Ltd. (Waltham, MA, USA). At the end of the experiment, the rats were sacrificed. Blood, liver, small intestine, and cecum contents were obtained and stored at –80°C. The liver was sectioned, preserved in formalin fixative, embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin and eosin (H&E).

Total RNA of liver and small intestine tissues was extracted using TRIzol^® Plus RNA Purification Kit following the manufacturer’s protocol. Single-stranded cDNA was synthesized with the Transcriptor First Strand cDNA Synthesis Kit. Real-time quantitative PCR (RT-PCR) was conducted to determine the expression of PPARα, cholesterol 7 alpha-hydroxylase A1 (CYP7A1), 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), adenosine monophosphate activated protein kinase (AMPK), ACC in the liver, carnitine palmitoyl transferase 1A (CPT1A), and ATP-binding cassette G5 (ABCG5), fatty acid translocase CD36 (CD36), Niemann-Pick C1-like 1 (NPC1L1), fatty acid transport protein 4 (FATP4), and microsomal triacylglycerol transfer protein (MTP) in the small intestine mRNA. Quantitative PCR primers were designed using Primer Premier 6.0 and Beacon Designer 7.8 and synthesized by Shenggong Bioengineering Co., Ltd. (Shanghai, China). RT-PCR was performed with PowerUp™ SYBRTM Green Master Mix (Applied Biosystems, Foster City, CA, USA) using CFX384 Real-Time Fluorescence Quantitative PCR System (Bio-Rad, Hercules, CA, USA). The primer sequences are shown in Table 1. The reaction conditions were as follows: 95°C for 1 min and 40 cycles of 95°C for 15 s and 63°C for 25 s to collect fluorescence. The expression of genes was normalized in reference to the housekeeping gene GAPDH, and the relative gene expression in the six groups was statistically analyzed using the 2^–ΔΔCt method.

Radio immunoprecipitation assay lysis. buffer including protease inhibitor cocktail (Thermo Fisher, Waltham, MA, USA) was used for the extraction of the total proteins of liver and small intestine, and the protein concentration was determined using BCA quantitative kit (Beyotime, Shanghai, China). The total extracted proteins were separated by SDS-PAGE for 2 h and then transferred to a PVDF membrane for 2 h. The membrane was incubated with T-TBS (containing 5% BSA) for 1 h and subsequently with the following primary antibodies: rabbit anti-CYP7A1 (1:1,000, Biorbyt orb539102, Cambridge, UK), anti-PPARα (1:1,000, Abcam ab126285, Cambridge, UK), anti-ABCG5 (1:1,000, Proteintech 27722-1-AP, Chicago, IL, USA), anti-NPC1L1 (1:500, Thermo Fisher PA5-72938, Waltham, MA, USA), anti-CPT1(1:500, Abcam ab128568, Cambridge, UK), and anti-GAPDH (1:10,000, Abcam ab181602, Cambridge, UK) as internal control at 4°C overnight. After being washed with T-TBS, the membranes were incubated with HRP-conjugated goat anti-mouse IgG secondary antibody (1:5,000, Thermo Pierce, Waltham, MA, USA) and goat anti-rabbit IgG secondary antibody (1:5,000, Thermo Pierce, Waltham, MA, USA) at room temperature for 1 h. Protein expression was visualized on X-ray films using SuperSignal^® West Dura Extended Duration Substrate (Thermo Pierce, Waltham, MA, USA). Image J 1.8.0 was used to analyze the optical density values of bands, and each test was repeated three times.

The cell fragments in the serum were removed by centrifugation at 10,000 rpm for 30 min, and the supernatant was transferred to Beckman L-100XP Ultracentrifuge (Brea, CA, USA) at 100,000 × g for 75 min and then discarded. The precipitate was washed and resuspended with PBS for another centrifugation at 100,000 × g for 75 min (24). After centrifugation, the supernatant was discarded, and the precipitate was resuspended with 200 μl of PBS to obtain serum extracellular vehicles (EVs). Quantification was performed using BCA kits (Thermo Fisher Co., Ltd., Waltham, MA, USA). In characterizing EVs, three or more proteins must be reported in at least a semi-quantitative manner, single vesicles must be examined, and the size distribution of EVs must be measured (25). EVs were characterized by TEM HT7700, and the average EV particle size was analyzed by Flow NanoAnalyzer N30E (Xiamen Fuliu Biological Technology Co., Ltd., Xiamen, China). EV-labeled proteins tetraspanins (CD9 and CD63) and endosome or membrane-binding proteins (TSG101) were quantitatively detected by Western blot.

For the selection of miRNAs closely related to the mRNAs of lipid metabolism, the species was first selected as rats on miRDB.¹ Several mRNAs related to lipid metabolism were then imported. One example is CYP7A1, that is, all miRNAs related to CYP7A1 gene were analyzed. For reliable results, all the miRNAs related to CYP7A1 gene obtained from the last website were inputted in another web site TargetScanMouse.² Hundreds of related genes were obtained for each miRNA. CYP7A1 was searched in this list. If the results overlap, then this miRNA will be selected as an alternative. Finally, miR-21-5p, miR-30b-5p, miR-33-5p, miR-27a-3p, and miR-126a-5p were identified as possibly interacting with target genes PPARα, MTP, CYP7A1, ABCG5, and HMGCR, respectively (14).

First, serum EV miRNA was extracted as follows. EVs were added with 300 μl of binding buffer, shaken evenly, and centrifuged at 12,000 g for 10 min. The supernatant was transferred to a spin cartridge, and the precipitate was retained and added with anhydrous ethanol to obtain a final ethanol concentration of 70%. The mixture was transferred to the second spin cartridge and centrifuged at 12,000 g for 1 min. The waste was discarded, and the sample was centrifuged again at 12,000 g for another 1 min after 500 μl of wash buffer was added to the spin cartridge. The waste liquid was discarded, and the above steps were repeated. Idling centrifugation (12,000 g, 5 min) was performed to dry the adsorption column, and centrifugation was conducted at 12,000 g for 2 min. Finally, the spin cartridge was added with 50 μl of RNase-free ddH₂O and then stored at –80°C after being placed at room temperature for 2 min (26). qRT-PCR was performed in line with the above steps. The reverse-transcription primer sequences are shown in Table 2, and the qRT-PCR primers are displayed in Table 3.

293T cells in logarithmic growth phase were seeded in 24-well plates and transfected with Lipofectamine™ 3000 Transfection Reagent at approximately 60% confluence. Before transfection, the medium was replaced with a fresh one. MiRNA-30b-5p mimics and miRNA-126a-5p mimics were prepared at 20 pmol/well, and recombinant plasmid pmirGLO-MTP-WT or pmirGLO-MTP-MUT and pmirGLO-HMGCR-WT or pmirGLO-HMGCR-MUT were added at 500 ng/well. The 24-well plates with fresh medium were placed in an incubator at 37°C with 5% CO₂ for 8 h. After 48 h, each well was washed with PBS twice and added with 250 μl of 1 × PLB lysate before the cells were lysed at room temperature. Approximately 100 μl of LAR II was absorbed in a black 96-well plate, followed by the addition of 20 μl of lysis solution. After the reading of mixture was checked, Stop & Glo substrate (100 μl) was added within 10 s and the new reading was recorded again.

Microbial DNA was extracted from rat cecal content samples with the Fast DNA SPIN kit (MPBIO, CA, USA) in accordance with the instructions. The V4–V5 region of the bacteria 16S ribosomal RNA gene was amplified by PCR under specific conditions (95°C for 2 min, followed by 25 cycles at 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s and a final extension at 72°C for 5 min) and using primers 515 F (5′-barcode-GTGCCAGCMGCCGCGG-3′) and 907 R (5′-CCGTCAATTCMTTTRAGTTT-3′); the barcode of which has an eight-base sequence unique to each sample. Amplicons were extracted from 2% agarose gels, purified with the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) and quantified using QuantiFluor™-ST (Promega, Madison, WI, USA). The Purified PCR products were quantified by Qubit^®3.0 (Life Invitrogen). The pooled DNA product was used to construct Illumina Pair-End library following Illumina’s genomic DNA library preparation steps. The amplicon library was then paired-end sequenced (2 × 250) on an Illumina Novaseq platform (Mingke Biotechnology Co., Ltd., Hangzhou, China). The image data files obtained by high-throughput sequencing were converted into sequencing reads, which were then stored in FASTQ files including their sequence information and sequencing quality information. For reliable results in subsequent analysis, the raw data were first filtered through Trimmomatic v0.33, CutAdapt 1.9.1 and custom Perl scripts were then used to identify and remove primer sequences, and high-quality reads without primer sequences were finally generated. USEARCH (version 10³) was employed to assign sequences with ≥ 97% similarity to the same OTUs. Tukey’s method and R software were used for the comparison of alpha diversity index between samples (P = 0.05). Beta diversity analysis and mapping were performed on QIIME (27) to reveal differences and similarities among the samples as measured by principal coordinate analysis (PCoA) with R software. LEfSe method was applied to compare the differential abundances of bacteria among groups at family and genus levels (28). Only those taxa with a log LDA score > 4 were considered.

Data were shown as mean ± standard deviation (SD). Statistical differences were analyzed by one-way ANOVA with Tukey’s multiple comparison test using SPSS 18.0 (P < 0.05 indicated significant differences).

The encapsulation rates of NF and EGCG by NFEG-microgel were 90 and 86%, respectively, and that of NF by liposomes was 93.5%. The average particle size, PDI, and zeta potential of NF-liposome were 104.8 nm, 0.226, and +59.8 mV, respectively, and those of NFEG-microgel were 2595 nm, 0.532, and –41.6 mV, respectively. Absolute zeta potentials greater than 30 were considered as stable. The microstructural morphologies of NFEG-microgel were observed by TEM as shown in Figure 1A. All the microgels had regular spherical or subspherical shape with neat edges. The average particle size of NFEG-microgel was 0.5∼3.0 μm (Figure 1B), which was in line with the above results. According to the rheological data, the shear viscosity of NFEG-microgel decreased with the increase in shear rate. The mathematical power-law function model was fitted to the obtained data using the cf tool in MATLAB R2020a to determine the consistency index k and fluidity index n. The fitting coefficient R² was 0.99, indicating that NFEG-microgel conforms to the power-law fluid, that is, it is a non-Newtonian fluid. The k value was 0.899, and the n value was 0.265. If n < 1 in power-law fluid, then NFEG-microgel is a fake plastic fluid (Figure 1D).

The liver of the rat in each group was photographed (Supplementary Figure 1). In terms of appearance, the liver of rat in the HFD + NF, HFD + EGCG, HFD + NFEG, HFD + MG, and HFD groups was larger and showed different degrees of yellowing compared with that in the normal group. The liver color of the HFD + NFEG group was the closest to that in the normal group. The body weights of different groups showed an increasing trend. At the end of HFD feeding for 8 weeks, the body weight of rats was significantly higher in the HFD group than that in the NC group but lower than that in the HFD + NF, HFD + EGCG, HFD + NFEG, and HFD + MG groups. The HFD + NFEG group showed the most remarkable reduction in body weight (Figure 2A). The total 8-week feed consumption in kilograms (kg) for the NC, HFD + NF, HFD + EGCG, HFD + NFEG, HFD + MG, and HFD groups was 5.67, 5.31, 5.04, 5.61, 5.51, and 5.69 kg, respectively. According to these data, the normal diet consumption per day was consistent with the rising trend of body weight gain in the NC group. A slight decrease in HFD consumption per day was observed in the other groups from the 5th week to the last week. The trends of water intake per day were in accordance with the diet consumption, and no difference in weekly water consumption was found among the groups.

As shown in Figure 2B, the NC, HFD + NF, HFD + EGCG, HFD + NFEG, HFD + MG groups had significantly lower serum TC, TG, and LDL-C concentrations (P < 0.05) and significantly higher HDL-C content (P < 0.05) than the HFD group, except for the HFD + MG group whose HDL-C content showed no significant difference from that in the HFD group. The difference between HFD + NFEG and HFD groups was the most significant (P < 0.01). In particular, the HFD + NFEG group had significantly lower TC and LDL-C contents than the HFD + NF group and significantly different TG and HDL-C contents compared with those in the HFD + NF and HFD + EGCG groups (P < 0.05).

Rat liver H&E staining results showed that the hepatocytes filled with large lipid composition had balloon-like changes, and the hepatocyte nuclei were constricted and deformed in the HFD group. Meanwhile, the amounts of lipid droplets and the degree of liver fat cavitation in the HFD + NF, HFD + EGCG, and HFD + NFEG groups were reduced compared with those in the HFD group, indicating that HFD-induced hepatic fat accumulation could be reduced by NF and EGCG (Figures 2C–H). These results demonstrated that NFEG-microgel has a better effect on improving the lipid profile, hepatic steatosis, and liver injury than HFD or feeding with single package NF or EGCG-microgel.

Key genes associated with lipid metabolism in the liver and small intestine were detected, and differences were observed among the groups. HFD + NFEG had significant effects on liver PPARα, AMPK, ACC, CYP7A1, CPT1A, and HMGCR (P < 0.01) compared with HFD. The expression of AMPK, CPT1A (P < 0.01), and CYP7A1 (P < 0.05) genes in the HFD + NFEG group were significantly higher than those in the HFD + NF and HFD + EGCG groups (Figure 3A). HFD + NFEG downregulated the expression of NPC1L1, MTP, CD36, and FATP4 and significantly repressed the downregulation of ABCG5 in the small intestine of HFD rats (P < 0.01). The HFD + NFEG group had significantly higher expression of ABCG5 gene (P < 0.01) and significantly lower expression of MTP genes (P < 0.05) compared with the HFD + NF and HFD + EGCG groups. FATP4 expression in the HFD + NFEG group was significantly decreased compared with that in the HFD + NF group (P < 0.01). CD36 and FATP4 expression showed no significant differences between the HFD + EGCG and HFD + NFEG groups, indicating that EGCG mainly inhibited lipid absorption after its release in the intestine. In addition, the expression of NPC1L1 and CD36 in the HFD + MG group was as low as that in the HFD + NFEG group (Figure 3B), indicating that the blank microgel had a similar effect on lipid absorption.

Western blot was used to detect liver PPARα, CYP7A1, HMGCR, and CPT1A and small intestine NPC1L1, MTP, and ABCG5. As shown in Figure 4, differences in the expression of these proteins were observed among the groups. The protein expression levels of PPARα, CYP7A1, and CPT1A were significantly higher in the HFD + NF, HFD + EGCG, and HFD + NFEG groups than in the HFD group (P < 0.01) and were higher in the HFD + NFEG group than in the HFD + NF and HFD + EGCG groups (P < 0.05). The protein expression level of HMGCR was significantly higher in the HFD and HFD + MG groups than other groups (P < 0.05) where that in the HFD + NFEG group was the lower than others (P < 0.05). The protein expression levels of NPC1L1 and MTP significantly decreased in the HFD + NF, HFD + EGCG, and HFD + NFEG groups compared with those in the HFD group (P < 0.01) and were lower in the HFD + NFEG group than in the NC, HFD + NF and HFD + EGCG groups (P < 0.05). The protein expression levels of ABCG5 were significantly higher in the HFD + NF, HFD + EGCG, and HFD + NFEG groups than in the HFD group (P < 0.01) and were higher in the HFD + NFEG group than in the HFD + NF and HFD + EGCG groups (P < 0.01). Meanwhile, the protein expression levels of PPARα, CPT1A, HMGCR, NPC1L1, and ABCG5 in the HFD + MG group showed no significant differences from those in the HFD group (P > 0.05), except for NPC1L1 and MTP (P < 0.05).

As shown in the TEM observation in Figure 5A, the serum EVs had spherical structures and saucer shape. Their particle size range within 50–140 nm (Figure 5B), and the average particle sizes of the NC, HFD + NF, HFD + EGCG, HFD + NFEG, HFD + MG, and HFD groups were 86.92, 83.71, 88.08, 96.01, 92.45, and 84.69 nm, respectively. For EV characterization, three marker proteins were reported in at least a semi-quantitative manner. Western blot analysis showed that the expression levels of the surface proteins of EVs, CD9, CD63, and TSG101 were high in the EVs (Figure 5C).

Five miRNAs closely associated with lipid metabolism were detected. Compared with those in the HFD and HFD + MG groups, the expression levels of miR-30b-5p and miR-126a-5p in the NC, HFD + NF, HFD + EGCG and HFD + NFEG groups were significantly increased (P < 0.05). No significant difference was observed in the expression level of miR-21-5p, miR-33-5p, and miR-27a-3p among the groups. In addition, the expression level of miR-30b-5p in the HFD + NFEG group was higher than that in the NC, HFD + NF, and HFD + EGCG groups (P < 0.05), and the expression level of miR-126a-5p showed no significant differences among the NC, HFD + NF, HFD + EGCG, and HFD + NFEG groups (Figure 5D). In accordance with the qRT-PCR results of lipid metabolism genes in the liver and small intestine, miR-126a-5p and miR-30b-5p may interact with their target genes HMGCR and MTP, respectively.

qRT-PCR results elucidated that the expression levels of miR-126a-5p and miR-30b-5p in the serum EVs significantly differed among the six treatment groups and showed negative correlation with the expression of HMGCR and MTP. TargetScan and miRDB databases predicted possible interaction sites between miR-126a-5p and miR-30b-5p and their target genes HMGCR and MTP mRNA, respectively. The interactions of miR-30b-5p-MTP and miR-126a-5p-HMGCR were further tested by constructing the reporter plasmids of the potential action points of HMGCR and MTP and conducting dual-luciferase reporter assay.

As shown in Figures 5E,F, the relative fluorescence value of rno-miR-126a-5p + HMGCR-WT co-transfection group were lower (P < 0.05) than that of miR-NC + HMGCR-WT co-transfection group. After HMGCR mRNA mutation, no difference in relative fluorescence value was observed between the co-transfections of miR-NC+HMGCR and rno-miR-126a-5p + HMGCR groups (P > 0.05; Figure 5E). Dual-luciferase reporter assay results of miR-30b-5p and MTP showed similar results (Figure 5F). The above results implied the existence of interaction sites for miR-126a-5p and miR-30b-5p and their respective target mRNAs, HMGCR and MTP. These sites may be the key to relieve NAFLD by post-transcription mechanism between miRNAs and mRNAs.

The gut microbiota may be an underlying target to treat obesity and related metabolic diseases. Bacterial 16S rRNA in stool was conducted to examine whether NFEG-microgel can alter the composition of the gut microbiota. Analysis of α-diversity index (Chao1 index, Shannon index, and Simpson index) reflecting gut microbiota showed that α-diversity increased with the increase in Chao1 index and Shannon index and the decrease in Simpson index (Figure 6A). Significant differences were observed among the HFD + NFEG and HFD, HFD + MG, and HFD + NF groups. Bray–Curtis PCoA analysis of OTU abundance of each rat revealed that compared with that in the NC group, the intestinal microbiota in the HFD group showed significant structural changes along the first principal component (PC1) and the second principal component (PC2). Compared with the NC group, the HFD + MG and HFD + EGCG groups showed the same shift as the HFD group along PC1 whereas the HFD + NFEG group had almost the same intestinal microbiota as the NC group along both PC1 and PC2 (Figure 6B). At the genus level, high-abundance species were selected to profile the expression of each group with the heatmap in Figure 6C. From a homoplastic viewpoint, linear discriminant analysis effect size (LEfSe) method was employed to identify statistically significant biomarkers and dominant microbiota among these groups (Figures 6D,E). Unexpectedly, Firmicutes was found to be the primary phylum of the gut microbiota in the HFD group, and this finding agreed with the study of Wang et al. (5). Meanwhile, Bacterioidetes was confirmed to be the dominating phylum of the gut microbiota in the HFD + MG group. The relative abundance of Allobaculum, Blautia, Bacteroides, Butyricicoccus, Phascolarctobacterium, Faecalibaculum, Ruminococcaceae UCG-013, and Turicibacter significantly increased (p < 0.05), and that of Romboutsia, Erysipelotrichaceae, and Lachnospiraceae_NK4A136_group significantly decreased (p < 0.01). No regulatory effect on Lactobacillus was found (Figures 7A–L). Experimental results showed that NF had a significant effect on Allobaculum (Figure 7A) and Erysipelotrichaceae (Figure 7J) because it only induced significant changes in the HFD + NFEG and HFD + NF groups. NFEG-microgel significantly enriched Bacteroides, Butyricicoccus, Phascolarctobacterium, Faecalibaculum, and Ruminococcaceae UCG-013 and significantly reduced Romboutsia (P < 0.01) as which shown in Figures 7B,D–G,I. The relative abundances of these genera in the HFD + NFEG were significantly different from those in the HFD + NF, HFD + EGCG, HFD + MG, and HFD groups (P < 0.05).