Pueraria lobata was purchased from a local market (Hangzhou, China). Then, the fruiting bodies of P. lobata root were grounded into powder and stored at a low-temperature refrigerator before being used. The antioxidant indicator kits of superoxide dismutase activities (SOD), glutathione peroxidase peroxide (GSH-Px) activities, total antioxidant capacity (T-AOC) activities, and malondialdehyde (MDA) contents were obtained from the Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The other chemicals and solvents were of liquid chromatography grade and purchased from Fisher Scientific or Beijing Solarbio Science & Technology Limited Company (Beijing, China).

The extraction method of P. lobata root was conducted according to a previous study with some modifications (16). In brief, P. lobata powder (10 g) was mixed with chilled acetone (200 ml) with ultra-sonicated for 30 min and then centrifuged for 20 min with the speed of 14,000 × g at 4°C. Then, the supernatant was removed, and the remaining pellet was re-extracted with another chilled acetone (200 ml). The pooled extracts were evaporated to dryness at 35°C. The macroporous resin column was used for the preliminary purification. The solution was then reconstituted in the solution of methanol, dried, and stored at −20°C before the test. The residue in this study referred to the P. lobata extract.

The total phenolic content of the P. lobata extract was measured based on the method described previously with some minor modifications (17). The P. lobata extract was dissolved in methanol to a concentration of 0.25 mg/ml. Gallic acid standards or the samples (20 μL) and diluted Folin–Ciocalteu reagent were incubated at room temperature for 1 min, followed by terminating reaction with sodium carbonate (75 mg/ml). Molecular Devices SpectraMax microplate spectrophotometer (San Jose, CA, United States) was used to record the absorbance at 760 nm. Total phenolic contents were expressed as μmol gallic acid equivalents/mg polyphenol-rich P. lobata extract.

The chromatographic analysis of the P. lobata extract was performed using a high-performance liquid chromatography (HPLC) system (Waters, Milford, MA, United States) (18, 19). The mobile phases were prepared as follows: water with 0.1% formic acid and acetonitrile with 0.1% formic acid. The gradient procedure was as follows: 0–3 min, 5% B, 3–42 min, 5%–35% B, 42–46 min, 35%–90% B, 46–50 min, 90% B, 50–56 min, 90%–5% B. The column was maintained at 30°C. The flow rate was set at 0.8 ml/min. The P. lobata extract was separated using an Agilent ZORBAX Eclipse Plus C18 (250 mm × 4.6 mm, 5.0 μm particle size). The detection of the P. lobata extract was recorded using an ultraviolet detector at 280 nm wavelength.

The animal experimental protocols were approved and authorized by the Institutional Animal Care and Use Committee of China Pharmaceutical University (SYXK-2020-0023). The experimental test was designed according to the study with minor modifications (22). Twelve specific pathogen-free male C57BL/6 mice (6 weeks of age, around 18 g) were purchased from the Model Animal Research Center of Nanjing University. The mice were divided randomly into two groups after acclimatization for 1 week. In brief, except for daily gavage with water, the two-group mice were separately fed by the normal diet with or without P. lobata extract (200 mg/kg/day). The body weight of each mouse was measured daily. Also, the feces were collected at the end of the feeding duration and the experimental period. The feces of two experimental groups were immediately frozen in liquid nitrogen and stored at −80°C before the sequence analysis. All the animals were sacrificed using isoflurane as an anesthetic. Livers from each mouse were excised and collected for biochemical assays.

The biochemical parameters of liver samples from the mice, including SOD, GSH-Px activity, T-AOC, and MDA activity, were prepared according to the study and determined by using commercial kits following the manufacturer’s instructions (23).

Total genome DNA was extracted from the fecal samples of the control and the P. lobata extract groups using a QIAamp DNA stool kit, and the DNA concentration and purity were monitored on 1% agarose gels, and then diluted to 1 ng/μl for amplicon generation. The 16S V3-V4 rDNA genes were amplified using the specific primer with the barcode. All PCR reactions were carried out with the high-fidelity PCR master mix, forward and reverse primers, and template DNA. Thermal cycles were conducted as the following: initial denaturation at 98°C for 1 min, 30 cycles of denaturation at 98°C for 10 s, annealing at 50°C for 30 s, and elongation at 72°C for 5 min. Then, the PCR products were mixed and purified using the Axygen AxyPrep DNA Gel Extraction Kit. Finally, the sequencing libraries were generated using the NEBNext^® Ultra™ DNA Library Prep Kit following the manufacturer’s instructions and then sequenced on an Illumina MiSeq platform (San Diego, CA, United States).

The paired-end reads from the original DNA fragments from the control and the P. lobata extract groups were merged using the Flash software. The aired-end reads were assigned by the different barcodes. Then, the Upare software package was used to perform the sequences. The alpha and beta diversities were also analyzed by the In-house Perl scripts. The sequences with ≥97% similarity were assigned to the same operational taxonomic units (OTUs). The observed species were used to estimate the amount of unique OTUs, as well as the rarefaction curves were generated based on these three metrics. The cluster analysis was illustrated using the principal component analysis and the principal coordinate analysis (PCoA). The Stamp software was utilized to study the differences in the abundances of individual taxonomy. The Linear discriminant analysis Effect Size (LEfSe) was used for the quantitative analysis of biomarkers among different groups. The 16S rRNA raw sequence data generated were deposited into the National Center for Biotechnology Information Sequence Read Archive Sprout and available under the accession number of PRJNA805070¹.

The SPSS software (version 24.0) and excel (version 2020) were used for statistical analysis. The Student’s t-test was used to assess the mean difference between the P. lobata extract and the control groups. A p-value of <0.05 or <0.01 was considered statistically significant. The results were expressed as the mean ± standard deviation.

Pueraria lobata has been in the Chinese List of Food and Medicine Homologous Substance since 2002 and is currently regarded as a functional food in the food industry due to possessing a great number of bioactive compounds such as isoflavones, polysaccharides, terpenoids, and amino acids. It has been reported that the antioxidant activity of the P. lobata extract was well correlated with the content of phenolic compounds (24). Although other antioxidants were probably present in the P. lobata extract, phenolic compounds made a significant contribution to the antioxidant activity. The polyphenol-rich P. lobata extract (EPM) was eluted by gradient elution HPLC (Figure 1). The yield content of this P. lobata extract is 1.2%. The total phenolic content of this extract was 67.3 mg gallic acid equivalents/g dry weight.

As the mechanisms of antioxidant evaluation methods in vitro and in vivo were different, in order to determine the potential antioxidative activities comprehensively, mice model was used, and the SOD, GSH-Px activity, T-AOC, and MDA contents in the mice liver were measured, which could be mirrored the cellular antioxidant status as key indicators (Table 1) (30). The main endogenous antioxidant enzymes including SOD, GSH-PX, T-AOC, and MDA in the body system were a protective mechanism, and these compounds could prevent oxidative stress and cell damage (31). Puerarin was reported to possess the potential antioxidative activities both in vitro and in vivo. The antioxidant effect in vivo was partly responded via increasing the activities of antioxidant enzymes, for instance, increasing GSH and the ratio of GSH/GSSG in rat liver, as well as reducing the related activity of caspase-3 relative to cell apoptosis (23).

There was an insignificant difference in the body weight between the control and the P. lobata extract groups during the feeding period. The GSH-Px activity in the liver in the P. lobata extract group was significantly increased by 21.6% compared with that in the control group (p < 0.05). As for the SOD activity, T-AOC and MDA parameters in the mice liver of the P. lobata extract group were significantly increased by 25.7, 13.6, and 33.9%, respectively (p < 0.05), compared with control. As a result, our study noted that the P. lobata extract could enhance the activities of antioxidant enzymes in the liver and inhibit the oxidative damage inside the body, which was evidenced for potentially affecting positively the protection of the liver. Similar findings have been reported by Senevirathne et al. (32). Polyphenols extracted from foods or medical herbs have been also demonstrated a strong ability to scavenge free radicals and terminate oxidative reactions, thereby in consideration of their antioxidant capacities (33, 34). Previously, dietary polyphenol-rich supplementation of barley bran and cranberry (Vaccinium macrocarpon) peel could help to boost the antioxidant status of serum and liver based on the mice model evaluation (35, 36). Notably, the results indicated that the enhanced antioxidant capacities of polyphenol-rich extract of P. lobata in vitro could be further verified in this mice liver model via increasing the activities of SOD, GSH-Px, and MDA, as well as did better in T-AOC than those in the control.

The P. lobata extract was proved to possess the antioxidant bioactivities using in vitro and in vivo methodologies, and to further investigate the potential effect on the gut microbes, the in vivo evaluation of the P. lobata extract on the composition and structure of the gut microbiota was performed (Figure 3A). Gut microbiota is responsible for the production of human functional compounds, vitamins, and metabolization of the dietary compound and has a potential beneficial protection against the unhealthy pathogens. Numerous studies showed that any imbalanced status about the composition and structure of gut microbiota would potentially cause the development of colonic dysbiosis, which is powerfully associated with several pathologies and human chronic diseases, including intestinal bowel disease, colon cancer, and obesity (37–39). Polyphenols derived from foods and medicine plants have the capability to reduce bacterial metabolizing enzymes, thereby reducing the colonic tumor incidence and improving cell metabolism; as well, the growth of harmful microbiome could be inhibited, and the abundance of the health-benefit microbiome was enhanced, such as Lactobacillus and Bifidobacterium species (40). The link of the antioxidant effect of polyphenols in vivo and hinder the bacterial proliferation, and the growth of enterpathogenic bacteria was also discussed in several studies (41, 42).

We performed the high-throughput sequencing of the 16S rDNA region to determine the gut microbial community treated by the P. lobata extract. The sequencing depth reads of the control and the P. lobata extract groups were in the range of 272,642 base pairs (bp) and 295,776 bp, respectively. The data indicated that most of the species in the gut microbiota were captured, and the output of these data was valid. Considering the bacterial change at the phylum level, the microbiome of the control and the P. lobata extract groups mainly consisted of Actinobacteria, Bacteroidetes, Cyanobacteria, Deferribacteres, Epsilonbacteraeota, Firmicutes, Patescibacteria, Proteobacteria, Tenericutes, and Verrucomicrobia (Figure 3B). The ratios of Actinobacteria and Bacteroidetes in the P. lobata extract group were 1.39-fold and 1.20-fold higher than those in the control group, respectively. However, the ratios of Cyanobacteria, Firmicutes, and Proteobacteria decreased by 15.5, 44.7, and 28.1% compared with the control group, respectively. The ratio of Bacteroidetes and Firmicutes phyla consisted of more than 90% population in the gut microbiota, and they would play an important role in human health (43). The Bacteroidetes was considered to have an effect to modulate the immune responses and activate the immune lymphocyte T cells (44). In summary, the supplementation of the P. lobata extract contributed to the functional differences of mice gut microbiota.

All the 10 most abundant or representative families were Muribaculaceae, Lachnospiraceae, Lactobacillaceae, Ruminococcaceae, Akkermansiaceae, Erysipelotrichaceae, Rikenellaceae, Bacteroidaceae, Prevotellaceae, and Burkholderiaceae (Figure 3C). The ratios of Lactobacillaceae, Rikenellaceae, and Bacteroidetes in the P. lobata extract group were 1.64-fold, 2.02-fold, and 3.96-fold higher than those in the control group, respectively. However, the ratio of Ruminococcaceae, Prevotellaceae, and Burkholderiaceae decreased by 66.8, 23.3, and 73.6% than those in the control group. Interestingly, the P. lobata extract almost decreased over 90% of Erysipelotrichaceae and Akkermansiaceae at the family level, which could be useful when these two families of microbiome were overgrown and overproduced in the gut. These results noted that the P. lobata extract would be a useful ingredient to restore the balance of the gut ecosystem.

When searching how the P. lobata extract and control differ in terms of the composition and structure of gut microbiota profiles, a distance square matrix was calculated to reflect the dissimilarity between the two groups (weighted UniFrac and unweighted UniFrac distances) (Figure 3D). The QIIME software could calculate both weighted and unweighted UniFrac distance, which are phylogenetic measures of beta diversity. We used unweighted UniFrac distance for the PCoA, and the PCoA could help to get principal coordinates and visualize them from complex, multidimensional data (Figure 3E). The results demonstrated that the bacterial communities of the P. lobata extract were significantly different from those of the control group. A long-distance between these groups was observed, pointing out that the P. lobata extract modulated the bacterial gut microbiome. This distance difference was also significantly evident in the non-metric multidimensional scaling, which displayed the differences among gut microbiota of the P. lobata extract and control groups (Figure 3F).

Altogether, the metagenomics showed the modulation of gut microbiota by the consumption of the P. lobata extract group via the changing of the abundance of Firmicutes and Bacteroidetes phyla and the decrease of Ruminococcaceae, Verrucomicrobiae, and Prevotellaceae phyla in mice gut (Figures 4A,B). Chen et al. administered 12.5 mg/kg polysaccharides extracted in P. lobata to both normal mice and antibiotic-associated diarrhea mice for 2 weeks (45). They showed that the P. lobata extract fractions significantly increased the abundance of beneficial bacteria, involving Oscillospira and Anaerotruncus, whereas relieving colonic pathological changes and gut microbiota dysbiosis additionally. Thus, P. lobata polyphenol extract may be a promising source of the phenolic compound that exerts prebiotic effects in vivo for human health. In addition to taxonomic composition, the functional profiles were predicted by PICRUSt, and the functional differences are shown in Figure 5. The LEfSe result could illustrate biological explanations to establish the biological consistency and estimation of predicted biomarkers (46). In the treatment of P. lobata, the phenolic-rich extracts induced functional differences of microbial communities consisted of cellular process and signaling, circulatory system, and neurodegenerative diseases. This was in consistent with the previous studies that the Puerarin, the main bioactive compound found in P. lobata, was found to be beneficial to the neurodegenerative diseases in vitro and in vivo (47, 48). A similar study found that the bioactive compound puerarin had an obvious anti-inflammatory effect using mice model, and it repaired intestinal mucosal integrity via regulating the abundance and kinds of the disorder gut microbiota, increasing the express of tight binding proteins ZO-1 and occludin, and these bacterial Ruminococcaceae, Lachnospiraceae, Eubacterium_xylanophilum, and Ruminococcus_1 were the key species related (49). This study provides new insight into the potential application of P. lobata to be developed as a functional ingredient, which would have a beneficial effect on gut health. The physiological effects of fermentation of P. lobata on the potential colon health will be further investigated.