Purslane polysaccharides (≥50%) are purchased from Lanzhou Wotelaisi Biotechnology Co., Ltd. (Wotls, Lanzhou, China). Methanol, distilled deionized water, formic acid, and acetonitrile (HPLC grade) were purchased from Thermo Fisher Scientific (Waltham, MA, United States); 3-NPH, EDC, and standard products (HPLC grade) were purchased from Sigma-Aldrich (St. Louis, MO, United States). Related analytical reagents were purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). The basic medium was prepared as the methodology reported by Mandalari et al. (2008) with minor modifications (2 g/l peptone, 2 g/l yeast extract, 0.1 g/l NaCl, 0.04 g/l K₂HPO₄, 0.04 g/l KH₂PO₄, 0.01 g/l MgSO₄⋅7H₂O, 0.01 g/l CaCl₂⋅6H₂O, 2 g/l NaHCO₃, 2 ml Tween 80, 0.02 g/l hemin, 10 μl vitamin K₁, 0.5 g/l cysteine HCl, and 0.5 g/l bile salts, pH 7.0).

Fecal samples collected by aseptic manipulation from five healthy Sprague-Dawley (SD) aged (21-month-old) rats were pooled with equal amounts and immediately placed inside an anaerobic chamber. 5 g of fecal samples was taken, and 20 times of the fecal weight of sterile saline was added to the anaerobic chamber, and the mixtures were fully shaken. The mixtures were filtered with eight layers of sterile gauze and transferred into another sterile sample bottle and then seeded into the basic medium to a final volume of 1 l for bacteria fermentation. The experiments were divided into six groups: Con1 (control group, fermented for 24 h), Con2 (control group, fermented for 48 h), LPOP1 (low-dose group of purslane polysaccharide fermented for 24 h), LPOP2 (low-dose group of purslane polysaccharide fermented for 48 h), HPOP1 (high-dose group of purslane polysaccharide fermented for 24 h), and HPOP2 (high-dose group of purslane polysaccharide fermented for 48 h). The fermentation volume of all groups was in 10 ml fermentation solution, the Con groups contained no POP, the LPOP groups contained POP (0.02 g/ml), and the HPOP groups contained POP (0.03 g/ml). Each group was repeated with five independent experiments. Gas production of the anaerobic cultures was recorded at 0, 12, 24, 36, and 48 h post inoculation (hpi), and the fermentation culture was collected at 24 and 48 hpi. Supernatants of the culture were separated by centrifugation for detection of pH, ammonia nitrogen, and short-chain fatty acid levels. The precipitates were stored at −80°C for further DNA extraction.

For ammonia analysis, 2.5 ml phenol chromogenic agent was added into a 10-ml tube, and 10 μl fermentation supernatants was added later. Then, 2.0 ml hypochlorite solution was also added into the tube. After shaking and mixing, the tube was incubated in water bath at 40°C for 15 min then cooled to room temperature to measure the absorbance value (OD₆₂₅), and the ammonia content was calculated according to the standard curve.

Short-chain fatty acids were analyzed by LC-MS/MS. A volume of 100 μl supernatant was added with 100 μl 50% acetonitrile-aqueous solution (v/v), and then grinding for 3 min. After that, the samples were sonicated in an ice water bath for 10 min and then centrifuged for 10 min, at 4°C, 12,000 rpm, followed by sample derivation and stored at −80°C for further analysis. The 5-μl samples were added into a high-performance liquid chromatograph (Nexera UHPLC LC-30A, Shimazin, Japan) with a color matching column (ACQUITY UPLC BEH C18 100 m × 2.1 mm × 1.7 μm, Waters) and a high-resolution mass spectrometer (AB SCIEX QTRAP 5500, AB SCIEX). The mass spectrum conditions are as follows: 0.1% formic acid-aqueous solution (A) and acetonitrile (B) are used as mobile phase, mass flow rate is 0.35 ml/min; curtain gas: 35 (psi); collision-activated dissociation (CAD) parameters: medium; negative ion spray voltage: −4,500 (V); ion source temperature: 450(°C); column temperature: 40(°C), ion source: gas 1:50 (psi); gas 2:60 (psi). Various short-chain fatty acids were determined according to the retention time, and the concentrations were calculated based on the standard curve.

Cecal content samples of the rats were collected and quick-frozen, then stored at −80°C for further analysis. Total genomic DNA was extracted using the DNA Extraction Kit (Magen, Guangzhou, China) according to the manufacturer’s instructions. The concentration of DNA was verified with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States) and agarose gel electrophoresis, respectively. All DNA samples were stored at −20°C for further study. PCR amplification of the V3–V4 hypervariable regions of the bacterial 16S rRNA gene was carried out with a 25-μl reaction mixture using universal primers (343F: 5′-TACGGRAGGCAGCAG-3′; 798R: 5′-AGGGTATCTAATCCT-3′). The reverse primer contained a sample barcode, and both primers were connected with an Illumina sequencing adapter.

The amplicon quality was visualized by gel electrophoresis. The PCR products were purified with Agencourt AMPure XP beads (Beckman Coulter Co., Brea, CA, United States) and quantified using the Qubit dsDNA assay kit. The concentrations were then adjusted for sequencing. Sequencing was performed on an Illumina NovaSeq 6000 with two paired-end read cycles of 250 bases each (Illumina Inc., San Diego, CA, United States; Oe Biotech Company, Shanghai, China).

Paired-end reads were preprocessed with Trimmomatic software (Bolger et al., 2014) to detect and cut off ambiguous bases (N). It also cut off low-quality sequences with an average quality score below 20 using a sliding window trimming approach. After trimming, paired-end reads were assembled using FLASH software (Reyon et al., 2012). Parameters of assembly were 10 bp of minimal overlapping, 200 bp of maximum overlapping, and 20% of maximum mismatch rate. Further denoising of sequences was performed as follows: reads with ambiguous, homologous sequences or below 200 bp were abandoned. Reads with 75% of bases above Q20 were retained by QIIME software (version 1.8.0) (Caporaso et al., 2010). After that, reads with chimera were detected and removed using VSEARCH (Rognes et al., 2016). Clean reads were subjected to primer sequence removal and clustering to generate operational taxonomic units (OTUs) using VSEARCH software with a 97% similarity cutoff (Rognes et al., 2016). The representative read of each OTU was selected using the QIIME package. All representative reads were annotated and blasted against Silva database (Version 132) using the RDP classifier (confidence threshold was 70%) (Wang et al., 2007). The microbial diversity in cecal content samples was estimated using the alpha diversity that includes Chao1 index, Shannon index, Observed-species, Simpson, Goods-coverage, and PD-whole-tree. The UniFrac distance matrix performed by QIIME software was used for unweighted UniFrac principal coordinate analysis (PCoA) and phylogenetic tree construction. The 16S rRNA gene amplicon sequencing and analysis were conducted by Oe Biotech Co., Ltd. (Shanghai, China).

Experimental data in the study are reported as mean ± standard deviation (SD) of at least biological duplicates. One-way analysis of variance (ANOVA) and Duncan’s multiple-range tests were employed to analyze significant differences in gas production, NH₃-N value, pH value, short-chain fatty acid (SCFA), and α-diversity (p < 0.05) between groups using SPSS version 19.0 (SPSS Inc., Chicago, IL, United States). Gas, NH₃-N, pH, and SCFA profiles were plotted in the GraphPad Prism version 8.0 (GraphPad Software, Inc., La Jolla, CA, United States).

As fermentation time increased, gas production in the HPOP and LPOP groups changed significantly compared with that in the Con group. Moreover, gas production in the HPOP group was significantly higher than in the LPOP and Con groups (Figure 1A). At the 24- and 48-h fermentation points, both ammonia nitrogen (NH₃-N) and pH values were significantly higher in the Con group than in the HPOP and LPOP groups (Figures 1B,C).

The results also showed that POP inhibited the production of short-chain fatty acids in the fermentation broth. After 24 and 48 h of fermentation, the concentrations of acetic acid (Figure 2A), butyric acid (Figure 2B), propionic acid (Figure 2C), and isovalerate (Figure 2D) in the fermentation broth of POP groups (both low- and high-dose groups) significantly decreased compared with Con groups; however, there was no difference between the LPOP and HPOP groups (Figure 2).

The Good’s coverage indices in the six groups were all greater than 99%, which indicated that most of the bacteria present in the samples were identified (Table 1). Purslane polysaccharides were found to significantly decrease fecal microbiota richness (Shannon, Simpson, and PD-whole-tree) and diversity (Chao1 and observed-species) after 24 or 48 h of in vitro fermentation compared to the control groups (Table 1). In addition, POP altered bacterial β-diversity (Figure 3). Principal component analysis (PCA) revealed the distinct clustering of microbiota compositions for the six groups. The microbiota compositions of LPOP1, LPOP2, HPOP1, and HPOP2 were not comparable; however, Con1 and Con2 were significantly comparable, but Con groups differed greatly from POP groups in the PC1 analysis, which was 16.1% of the total variation (Figure 3).

Microflora in fermentation cultures of different groups are mainly composed of Firmicutes, Bacteroidetes, Proteobacteria, and Actinobacteriota when categorized at the phylum level. Nine and six different phyla were observed with ANOVA analysis after 24 and 48 h of fermentation, respectively. Compared with the Con1 group, the abundance of Firmicutes in the LPOP1 and HPOP1 groups was significantly upregulated, while Bacteroidetes, Proteobacteria, Fusobacteriota, Desulfobacterota, Actinobacteriota, Acidobacteriota, Cyanobacteria, and others showed a significant decrease (Figure 4A). In addition, compared with the Con2 group, the relative abundance of Firmicutes in the LPOP2 and HPOP2 groups was significantly upregulated, whereas the relative abundance of Proteobacteria, Desulfobacterota, Bacteroidota, Fusobacteriota, and others was significantly decreased (Figure 5A).

At the genus level, Escherichia-Shigella and Lactobacillus were dominant in the Con and POP groups, respectively. After fermentation for 24 and 48 h, ANOVA analysis showed that 87 and 78 different bacterial genera were observed, respectively. Compared with the Con1 group, the relative abundance of Lactobacillus in the LPOP1 and HPOP1 groups increased significantly. However, other genera of microflora such as Escherichia-Shigella, Parasutterella, and Klebsiella decreased significantly (Figure 4B). Compared with the Con2 group, the relative abundance of Lactobacillus in the LPOP2 and HPOP2 groups was much higher. Conversely, the relative abundance of Escherichia-Shigella, Parasutterella, Phascolarctobacterium, Eubacterium nodatum_group, and Klebsiella decreased significantly (Figure 5B).

To identify specific microflora associated with POP, the linear discriminant analysis effect size (LEfSe) method was used to compare the differences in microflora between the Con and POP-treated groups after fermentation. The cladding diagrams of fermentative microorganisms and dominant bacteria composition are presented in Figures 6B, 7B. The most differentially abundant taxa are shown in Figures 6A, 7A. After 24 h of fermentation, the relative abundance of Bacilli–Lactobacillus, Eggerthella, and Firmicutes in the LPOP1 group was significantly increased, and in the HPOP1 group, the relative abundance of other bacteria was also significantly increased. In addition, the Escherichia–Shigella, Bacteroides, Clostridium sensu stricto 13, Enterobacter, Intestinimonas, Gardnerella, Proteus, Ruminococcus torques group, Eubacterium_ nodatum group, Prevotella, Clostridium sensu stricto 18, Caldibacillus, and Eisenbergiella in the Con1 group showed a significant increasing trend (Figure 6A). Furthermore, when the fermentation time was extended to 48 h, compared with the Con group, the relative abundance of Bacilli–Lactobacillus and Firmicutes in the LPOP2 group and Paraprevotella in the HPOP2 group showed a significant increasing trend. In the Con2 group, the relative abundance of several other phyla (i.e., Escherichia–Shigella, Bacteroides, Parasutterella, Enterobacter, Veillonella, Clostridium sensu stricto 13, Marinobacter, Parabacteroides, Erysipelotrichaceae UCG-003, Ruminococcus_torques_group, Tyzzerella, Mobiluncus, Gardnerella, and Clostridium sensu stricto 18) also significantly increased (Figure 7A).

Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) is designed to predict metagenome functional content from marker gene surveys and full genomes. The marker gene of 16s rRNA was used to amplify the subsequence to estimate the metabolomic function spectrum of a microbial community. According to the PICRUSt results, there were significant differences in metabolic potential between the Con and POP groups (Figure 8).

Compared with the POP-treated group, the Con groups showed an increasing trend in the expression of genes associated with environmental adaptation, xenobiotics biodegradation and metabolism, metabolism of cofactors and vitamins, metabolism of terpenoids and polyketides, lipid metabolism, metabolism of amino acids, carbohydrate metabolism, glycan biosynthesis and metabolism, energy metabolism, folding, sorting and degradation, transport and catabolism, cellular processes, signaling and metabolism, membrane transport, cell motility, biosynthesis of other secondary metabolites, cancers, metabolic diseases, cardiovascular diseases, neurodegenerative diseases, infectious diseases, circulatory system, digestive system, immune system, endocrine system, enzyme families, poorly characterized transcription, signal transduction, and genetic information processing. Conversely, the expression of genes involved in cell growth and death, translation, immune system diseases, signaling molecules, and interaction decreased.

Regarding fermentation time, in both LPOP1 and LPOP2 groups, genes of replication and repair, nucleotide metabolism, excretory system, nervous system, cell growth and death, translation, immune system diseases, xenobiotics biodegradation and metabolism, metabolism of cofactors, cell communication, and sensory system showed a noticeable change in expression level. The expression of genes for environmental adaptation, replication, and repair in the HPOP1 and HPOP2 groups was also significantly changed.