Modulation of Human Gut Microbiota In Vitro by Inulin-Type Fructan from Codonopsis pilosula Roots

Abstract

Inulin-type fructan (ITF) defined as a polydisperse carbohydrate consisting mainly of β-(2–1) fructosyl-fructose links exerts potential prebiotics properties by selectively stimulating the growth of Bifidobacterium and Lactobacillus. This study reported the modulation of human gut microbiota in vitro by ITF from Codonopsis pilosula roots using 16S ribosomal RNA gene sequencing. The microbiota community structure analysis at genus levels showed that 50 mg/mL ITF significantly stimulated the growth of Prevotella and Faecalibacterium. LEfSe analysis showed that ITF at 25 and 50 mg/mL primarily increased the relative abundance of genera Parabacteroides and Alistipes (LDA Score > 4), and genera Prevotella and Faecalibacterium (LDA Score > 4) as well as Acidaminococcus, Megasphaera, Bifidobacterium and Megamonas (LDA Score > 3.5), respectively. Meanwhile, ITF at 25 and 50 mg/mL exhibited the effects of lowering pH values of samples after 24 h fermentation (p < 0.05). The results indicated that ITF likely has potential in stimulating the growth of Prevotella and Faecalibacterium as well as Bifidobacterium of human gut microbiota.

Introduction

Inulin-type fructan (ITF), a kind of indigestible carbohydrate, exhibits a linear structure comprised of β-D-fructose via β-(1 → 2)-fructosyl glycosidic bonds and terminated with an α-D-glucose. Increasing evidence showed that ITF exhibited multiple bioactivities such as neuroprotection, immunoregulation, gut microbiota regulation, prebiotic activity, and some potential pharmaceutical functions including stabilization of proteins and modified drug delivery [1–5]. In particular, the prebiotic activity of ITF especially stimulating Bifidobacterium and Lactobacillus bacteria, is frequently highlighted [6–8].

Degrees of polymerization (DP) are commonly used to represent the number of fructose units in ITF. Different ITFs from different plant sources exhibit different DP. Generally, the short-chain ITF exhibits a DP of 2–10 and the long chain ITF refers to those with a DP of 10–60 [8]. The modulatory effect of ITF on gut microbiota is affected by both the DP of ITF and the host gut bacteria ecology.

Radix Codonopsis, derived from the roots of C. pilosula (Franch.) Nannf., C. pilosula (Franch.) Nannf. Var. modesta (Nannf.) L.T. Shen and C tangshen Oliv, has been used as traditional Chinese medicine beneficial for spleen and lung in China [9]. In recent years, C. pilosula polysaccharides have attracted more and more attention of researchers and have been reported to show various functions. Gong et al. found that C. pilosula polysaccharides had a protective effect on the lungs in LPS- and Escherichia coli-induced acute lung injury mouse models by attenuating the activation of mitogen-activated protein kinase (MAPK) and nuclear factor kappa B (NF-κB) induced by LPS and E. coli [10]. Wang and his colleagues reported that C. pilosula polysaccharides exhibited neuroprotective effects by attenuating Tau hyperphosphorylation and cognitive impairment in hTau infected mice and inhibiting Aβ toxicity and cognitive defects in APP/PS1 mice [11, 12]. A polysaccharide isolated from C. pilosula roots attenuated carbon tetrachloride-induced liver fibrosis by modulating TLR4/NF-κ and TGF-β1/Smad3 signaling pathway [13]. Xie et al. showed that oligosaccharides and polysaccharides from C. pilosula exhibited antifatigue and antihypoxia activities in mice basing on the weight-loaded swimming test and the normobarie hypoxia test, respectively [14].

ITF is one of polysaccharides in C. pilosula roots and frequently isolated by many researchers. As early as 2005, Ye et al. reported the isolation of an ITF from C. pilosula and its antitumor activity in vitro [15]. Recently, increasing researches regarding ITF from C. pilosula and its bioactivities such as anti-inflammatory and prebiotic activities are arising [16, 17]. Previously, we obtained an ITF with an estimated DP of 31 from C. pilosula roots, and it exhibited obvious anti-gastric ulcer effects [18]. In this study, we reported the isolation and human gut bacteria modulation in vitro of an ITF with an estimated DP of 13 from C. pilosula roots. The results demonstrated that the ITF exhibited significant growth stimulation on genera Prevotella and Faecalibacterium, as well as Bifidobacterium at a higher fermented concentration in in vitro fecal fermentation.

Materials and Methods

Materials and Chemicals

The roots of C. pilosula (Franch.) Nannf. were collected from Pingshun county in Shanxi province, China, and identified by Professor Jianping Gao. The specimen (2020–10-MCM01) was kept at the School of Pharmaceutical Science, Shanxi Medical University. D101 macroporous resin (AR) was purchased from KeLong Chemical Reagent Factory (Chengdu, China) and 95% ethanol was purchased from Hebei RuiKang Pharmaceutical Technology Co., Ltd. (Hengshui,China).

Extraction and Isolation of ITF from C. pilosula Roots

The dried roots of C. pilosula (5 kg) were crushed and extracted two times with methanol (10:1, v/w) by reflux at 60 °C for 1 h. The residue after filtration with gauze was extracted two times with distilled water (10:1, v/w) by reflux at 100 °C for 2 h. The water extract was subjected to D101 macroporous resin column chromatography eluted with distilled water. The water elution was precipitated with 80% ethanol to obtain the precipitate (270 g). The precipitate was dissolved in distilled water and subjected to a 5,000 ultrafiltration membrane (Millipore Corporation, US) to obtain ITF (180 g) after freeze-dried.

Homogeneity and NMR Analysis of ITF

The homogeneity of ITF was determined by high-performance gel permeation chromatography (HPGPC) performed with a LC-10AT HPLC system (SHIMADZU, Kytot, Japan) fitted with a Tskgel G4000 PWXL column (300 mm × 8 mm) and a Shodex RI-20H refractive index detector. The mobile phase was ultrapure water and the flow rate was 0.3 mL/min at 35 °C. ITF (30 mg) was dissolved in D₂O in a 5-mm NMR tube and measured for ¹H- and ¹³C-NMR on an AVANCE III 400 MHz spectrometer (Bruker, Rheinstetten, Germany).

Fecal Fermentation In Vitro

The fresh feces samples were collected from three healthy donors (Chinese, 1 Male and 2 Female, aged 23–28) who had not taken any antibiotics for at least six months, which was approved by the Research Ethics Committee of Shanxi Medical University (2022SJL25). Then the fecal samples were mixed with an equal amount of feces (67 g) from each donor, and diluted with 1000 mL PBS (pH 7.0) to obtain a 20% (w/w) solution, and subsequently mixed well using a vortex mixer. Then the feces samples were filtered through three layers of soup bags to remove food-derived debris and immediately transferred into an anaerobic workstation (A55, Don Whitley Scientific, UK). The atmosphere composition in the anaerobic workstation was 85% N₂, 5% CO₂, and 10% H₂.

The basal medium was prepared according to the previous method [19] with minor modifications. Generally, 0.5 g peptone, 1.0 g yeast extract, 25.0 mg NaCl, 10.0 mg K₂HPO₄, 10.0 mg KH₂PO₄, 2.5 mg MgSO₄, 2.5 mg CaCl₂, 0.5 g NaHCO₃, 5.0 mg hemin, 115.0 mg cysteine-HCl, 125.0 mg bile salt, 0.25 g resazurin, 0.5 mL Tween 80 and 2.5 µL vitamin K1 were dissolved in 250 mL distilled and the pH was adjusted to 7.0 with 0.1 mol/L HCl solution. The basal medium was sterilized at 121 °C for 20 min before use.

A volume of 5 mL of sterilized ITF solution at a designed concentration (25 mg/ml and 50 mg/ml), ultrapure water (CK group) was thoroughly mixed with 22.5 mL of fecal inoculum and 22.5 mL of basal medium using a vortex mixer. The experimental samples were transferred into different anaerobic sealed tubes, which were subsequently incubated at 37 °C and cultured for 24 h. Five replicate experiments were performed under the same conditions.

Analysis of Gut Microbiota

After fermentation for 24 h, PureLink™ Microbiome DNA purification kit (Invitrogen, USA) was used for extracting DNA from stool samples according to the manufacturer's instructions. The PCR amplification of the V3-V4 hyper variable region of the bacterial 16S rRNA was done with universal primers: 338 F (5′-ACTCCTACGGGAGG CAGCAG-3′), and 806R (5′- GGACTACHVGGGTWTCTAAT-3′). Next-generation sequencing of the V3–4 region of the bacterial 16S ribosomal RNA gene was performed on NovaSeq 6000 (Illumina, Novogene). Sequenced DNA was analyzed using QIIME2 (version 2019.7), and alpha-rarefaction curves were achieved. Taxonomy was determined using the Greengenes classifier. A phylogenetic tree was generated and Beta diversity was calculated using Unweighted UniFrac distance matrices. Alpha diversity was calculated using Faith’s Phylogenetic Diversity test. An unweighted UniFrac distances dot-plot was created using Prism (GraphPad, USA), version 8.2.1.

Determination of pH Value

The fecal slurries supplemented with ITF or water were sampled after 24 h and plunged into ice water for 5 min. These fecal slurries were centrifuged at 10,000 rpm for 10 min to separate the supernatants (fermentation medium), and then the pH value of each supernatant was measured.

Statistical Analysis

The normalized taxonomic profile and Alpha, and Beta diversity indices were used for statistical tests by using the R software (http://www.r-project.org/). Differences between any of the two groups were analyzed using Wilcoxon's test. To ascertain whether the visually observed differences in microbiota community among groups were statistically significant, adonis analyses were performed. A p value lower than 0.05 was considered statistically significant.

Results and Discussion

Homogeneity and NMR Analysis of ITF

HPGPC was a method widely used for detecting the homogeneity of a polysaccharide. As shown in Fig. 1A, ITF exhibited a single and symmetrical sharp peak at 34.382 min on the HPGPC chromatogram, indicating that ITF was a homogeneous polysaccharide.

Fig. 1.

HPGPC (A), ¹H- (B) and ¹³C-NMR (C) spectra of ITF

¹H-NMR (Fig. 1B) and ¹³C-NMR (Fig. 1C) of ITF were completely consistent with the published data for ITF [7, 17, 20]. Briefly, in the ¹H-NMR spectrum of ITF, several signals from 3.50 to 5.50 ppm occurred. The signal at 5.33 ppm was assigned to H-1 of terminal glucose residues in ITF and the signals at 4.15 and 4.00 ppm were assigned to H-3 and H-4 of fructose residues, respectively. The integration ratio of H3-Fru to H1-Glu was frequently used to estimate the DP value of ITF [7]. So, the DP of ITF was estimated to be 13 in this study as shown in Fig. 1B. In the ¹³C-NMR, six obvious carbon signals for fructose residues in ITF occurred at 103.18 ppm (C-2), 81.01 ppm (C-5), 76.91 ppm (C-3), 74.20 ppm (C-4), 62.06 ppm (C-6) and 60.82 ppm (C-1). Also, the signal at 92.42 ppm was assigned to C-1 of terminal glucose residues in ITF [7, 17, 20].

α-Diversity and β-Diversity of Human Gut Microbiota

The α-diversity index Shannon diversity (Fig. 2B) indicated that there were no significant differences between the CK group and ITF groups. Principle co-ordinates analysis (PCoA) is one of the important analytical methods for β-diversity. As shown in Fig. 2C, PCoA1 and PCoA2 contributed 59.52% and 15.03% of the variation, respectively. These results demonstrated that the gut microbiota in the ITF groups were significantly different from that in the CK group.

Fig. 2.

Structure of ITF (A), Shannon Diversity (B) and PCoA (C) of gut microbiota of different group fecal samples

Composition of Human Gut Microbiota

The microbiota community structure of fecal samples was examined at the phylum and genus levels. As shown in Fig. 3A, Bacteroidota, Firmicutes, Proteobacteria, Verrucomicrobiota, and Fusobacteriota were the dominant bacteria in all fecal samples at the phylum level, and the dominant bacteria Bacteroidota and Firmicutes accounted for 88.8% and above 94% of total bacteria in CK group and in ITF groups, respectively. The ratio of Bacteroidota to Firmicutes between groups exhibited no significant differences. It was interesting that the relative abundance of Fusobacteriota was significantly lowered in ITF groups compared to the CK group, which was consistent with the previous study showing that polysaccharides from loquat leaves and fructooligosaccharide (FOS) decreased significantly the relative abundance of Fusobacteria at the phylum level in an in vitro fecal fermentation [21]. The phylum Fusobacteriota occasionally exhibited potential pathogenicity. Qi et al. reported that the relative abundance of Fusobacteriota at the phylum level was much higher in Tibetan piglets with yellow dysentery than that in Tibetan piglets with white dysentery and healthy Tibetan piglets [22]. In cyclophosphamide-induced immunosuppressed rats, it was found that the bacterial in rat feces were enriched in Proteobacteria, Fusobacteriota and Actinobacteriota at the phylum level, which was then eliminated by the treatment of a combination of live Bifidobacterium, Lactobacillus, Enterococcus, and Bacillus [23].

Fig. 3.

The microbiota community structure at the phylum level (A) and the genus level (B) of the different group fecal samples

At the genus level (Fig. 3B), it was shown that the dominant bacterium was Bacteroides (Bacteroidetes), Alistipes (Bacteroidetes), Prevotella (Bacteroidetes), Faecalibacterium (Firmicutes) and Parabacterium (Bacteroidetes) in all fecal samples, and Bacteroides was the most dominant genus, which was consistent with the report that the bacterial phyla Bacteroidetes and Firmicutes were dominated in gut of most healthy adults, and the genus Bacteroides was one of the most dominant genera in Bacteroides phylum [24]. The relative abundances of Prevotella and Faecalibacterium were significantly increased in the ITF (50 mg/mL) group compared to the CK group. The genus Prevotella was found to exhibit the capacity to break down a wide-range of plant-derived polysaccharides and was associated positively with high levels of carbohydrate, fruit and vegetable intake [25]. Prevotella histicola, an emerging probiotic, significantly alleviated depression in ovariectomized mice by improving the intestinal microbiota to inhibit central inflammation and may be therapeutically beneficial for postmenopausal depression [26]. The study by Kovatcheva-Datchary indicated that dietary fiber consumption increased the abundance of Prevotella in gut microbiota and improved glucose metabolism potentially by increasing glycogen storage [27]. So, ITF of this study was a potential prebiotic by stimulating Prevotella.

The genus Faecalibacterium is prevalent in the human gut and is one of the most promising microbes for the development of probiotics and biotherapeutics, and F. Prausnitzii, the species of the genus Faecalibacterium, is functionally considered a significant biomarker of gut health [28, 29]. The critical role of F. Prausnitzii in human health was reviewed in detail, which presented that the decrease of abundance in F. Prausnitzii was associated with gastrointestinal disorders such as inflammatory bowel syndrome (IBS), inflammatory bowel disease (IBD) and colorectal cancer, and with metabolic syndromes such as obesity, type 2 diabetes, non-alcoholic fatty liver disease, and other diseases [30]. These data demonstrated that ITF could exhibit health benefits by increasing the relative abundances of Prevotella and Faecalibacterium in gut microbiota.

LEfSe Analysis

LEfSe (LDA Effect Size) is a useful analytical tool to find biomarkers with statistical differences among groups. As shown in Fig. 4, ITF at 25 mg/mL primarily increased the relative abundance of genera including Parabacteroides and Alistipes (LDA Score > 4) (Fig. 4A), and ITF at 50 mg/mL significantly increased the relative abundance of genera Prevotella, Faecalibacterium, (LDA Score > 4) as well as Acidaminococcus, Megasphaera, Bifidobacterium and Megamonas (LDA Score > 3.5) (Fig. 4B). The genus Alistipes consists of thirteen species, among of which some species exhibit helpful effects on human health, such as protecting against liver fibrosis and colitis, and some species are pathogenic in colorectal cancer [31]. The genus Parabacteroides also exhibits both beneficial and pathogenic effects on human health, such as protecting against colorectal cancer and obesity on the one hand and promoting inflammatory bowel disease on the other hand [32]. Meanwhile, ITF at both 25 mg/mL and 50 mg/mL significantly decreased the abundance of Fusobacterium compared to the CK (LDA Score > 4), which was consistent with the report by Yu et al. that inulin exhibited significant inhibition on pathogenic bacteria Fusobacterium [33].

Fig. 4.

LEfSE results among different groups (A and B) and the effect of ITF on pH values (C). ** p < 0.05 compared to CK

These data indicated that ITF exhibited different impacts on human gut microbiota at different concentrations, which reminded us that the dose was an important factor to be considered in evaluating the effect of some substances on the composition of gut microbiota.

Effects of ITF on pH Value

Short-chain fatty acids (SCFAs) mainly including acetate, propionate, and butyrate, are the main products from the fermentation of dietary fiber in the colon, and are responsible for the drop in pH in fermentation progress [34, 35]. The genus Prevotella was well known to produce succinate as a major product of fermentation in utilizing polysaccharides, which was rapidly decarboxylated to produce propionate playing a key role in gut protection [36]. The genus Faecalibacterium was also well-known butyrate-producing probiotics [37, 38]. As shown in Fig. 4C, the values of pH were significantly decreased after fermentation by ITF compared to the CK group (p < 0.05). The reason for lowering pH of ITF is not clear, and whether it is associated with production of SCFAs or not need to be revealed in subsequent experiments.

In conclusion, ITF is usually convinced to be prebiotic stimulating the growth of Bifidobacterium or Lactobacillus. However, ITF in this study exhibit the function of promoting the growth of Prevotella, Faecalibacterium, as well as Bifidobacterium and others and inhibiting the growth of Fusobacterium and others in human fecal fermentation in vitro.

Declarations

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.