In Vitro Fermentation of Pleurotus eryngii Mushrooms by Human Fecal Microbiota: Metataxonomic Analysis and Metabolomic Profiling of Fermentation Products

Abstract

Edible mushrooms contain biologically active compounds with antioxidant, antimicrobial, immunomodulatory and anticancer properties. The link between their anticancer and immunomodulatory properties with their possible prebiotic activity on gut micro-organisms has been the subject of intense research over the last decade. Lyophilized Pleurotus eryngii (PE) mushrooms, selected due to their strong lactogenic effect and anti-genotoxic, immunomodulatory properties, underwent in vitro static batch fermentation for 24 h by fecal microbiota from eight elderly apparently healthy volunteers (>65 years old). The fermentation-induced changes in fecal microbiota communities were examined using Next Generation Sequencing of the hypervariable regions of the 16S rRNA gene. Primary processing and analysis were conducted using the Ion Reporter Suite. Changes in the global metabolic profile were assessed by 1H NMR spectroscopy, and metabolites were assigned by 2D NMR spectroscopy and the MetaboMiner platform. PLS-DA analysis of both metataxonomic and metabolomic data showed a significant cluster separation of PE fermented samples relative to controls. DEseq2 analysis showed that the abundance of families such as Lactobacillaceae and Bifidobacteriaceae were increased in PE samples. Accordingly, in metabolomics, more than twenty metabolites including SCFAs, essential amino acids, and neurotransmitters discriminate PE samples from the respective controls, further validating the metataxonomic findings.

Keywords: Pleurotus eryngii mushrooms; gut microbiota; in vitro static batch fermentation; metabolomics; metataxonomics.

Figure 1

Box plots of Chao1 (a,c) and Shannon (b,d) alpha diversity measure estimates at family (a,b) and genus (c,d) levels. The X indicates the mean value; the circles indicate the samples’ alpha diversities; ‘*’ indicates a statistically significant difference as measured by paired-samples t-test with a p-value of < 0.05. NC0: Samples before fermentation in the absence of an additional carbon source (negative controls at 0 h), NC24: Samples after 24 h of fermentation in the absence of an additional carbon source (negative controls at 24 h), PE0: Samples before fermentation in the presence of lyophilized mushroom powder of P. eryngii, PE24: Samples after 24 h of fermentation in the presence of lyophilized mushroom powder of P. eryngii.

Figure 2

Metataxonomic analysis of families’ abundance in NC and PE in vitro samples both before and after fermentation. (A) Partial Least Square Regression-Discriminant Analysis (PLS-DA) scores plot (R2 = 0.69, Q2 = 0.40, accuracy = 0.46), (B) VIP scores plot of important families identified by PLS-DA. The colored boxes on the right indicate the relative concentrations of the corresponding family in each group under study, (C) PLS-DA model validation by permutation tests based on separation distance. The p value based on permutation is p < 0.001 NC0: Samples before fermentation in the absence of an additional carbon source (negative controls at 0 h), NC24: Samples after 24 h of fermentation in the absence of an additional carbon source (negative controls at 24 h), PE0: Samples before fermentation in the presence of lyophilized mushroom powder of P. eryngii, PE24: Samples after 24 h of fermentation in the presence of lyophilized mushroom powder of P. eryngii.

Figure 3

Metataxonomic analysis of genera abundance in NC and PE in vitro samples both before and after fermentation. (A) Partial Least Square Regression-Discriminant Analysis (PLS-DA) scores plot (R2 = 0.87, Q2 = 0.42, accuracy = 0.48), (B) VIP scores plot of important genera identified by PLS-DA. The colored boxes on the right indicate the relative concentrations of the corresponding family in each group under study, (C) PLS-DA model validation by permutation tests based on separation distance. The p value based on permutation is p < 0.014. NC0: Samples before fermentation in the absence of an additional carbon source (negative controls at 0 h), NC24: Samples after 24 h of fermentation in the absence of an additional carbon source (negative controls at 24 h), PE0: Samples before fermentation in the presence of lyophilized mushroom powder of P. eryngii, PE24: Samples after 24 h of fermentation in the presence of lyophilized mushroom powder of P. eryngii. The Square brackets ([]) around a genus indicates that the name awaits appropriate action by the research community to be transferred to another genus.

Figure 4

Volcano plots for the comparison of microbiota abundance among fecal samples that underwent 24 h of fermentation with and without the presence of P. eryngii. (A) families, (B) genera. Estimate = log2 fold change; the vertical red line is set to Estimate = 0; the horizontal red line is set to FDR = 0.05.

Figure 5

Volcano plots for the comparison of microbiota abundance among fecal samples without the presence of P. eryngii and with or without fermentation (NC0 and NC24, respectively). (A) families, (B) genera. Estimate = log2 fold change; the vertical red line is set to Estimate = 0); the horizontal red line is set to FDR = 0.05.

Figure 6

PLS–DA analysis for the pre- and post-fermentation samples from eight volunteers. (A) Scores plot of PLS-DA analysis (R2X(cum) = 0.90, Q2(cum) = 0.81, accuracy = 0.97), (B) VIPs plot of the studied metabolites. NC0: Samples before fermentation in the absence of an additional carbon source (negative controls at 0 h), NC24: Samples after 24 h of fermentation in the absence of an additional carbon source (negative controls at 24 h), PE0: Samples before fermentation in the presence of lyophilized mushroom powder of P. eryngii, PE24: Samples after 24 h of fermentation in the presence of lyophilized mushroom powder of P. eryngii, (C) Validation of the PLS–DA analysis, by permutation test statistics, indicating that the extracted model is significantly different from a model built on random data. The permutation tests were carried out with 1000 random permutations, thus providing significance of the model at the 0.001 level.

Figure 7

Volcano plots illustrating the changes in metabolite concentrations caused by the P. eryngii addition into the pre-fermentation samples (PE0 vs. NC0). Estimate = log2 fold change; the vertical red lines are indicative and are set to fold change 1.5); the horizontal red line is set to FDR = 0.05.

Figure 8

PLS–DA analysis of post- fermentation samples from eight volunteers. (A) Scores plot of PLS-DA analysis (R2X(cum) = 0.90, Q2(cum) = 0.69, accuracy = 1.00). (B) VIPs plot of the studied metabolites. NC24: Samples after 24 h of fermentation in the absence of an additional carbon source (negative controls at 24 h), PE24: Samples after 24 h of fermentation in the presence of lyophilized mushroom powder of P. eryngii.

Figure 9

Volcano plots for the pairwise (paired Wilcoxon test) differential metabolic composition of fermented samples with or without the addition of P. eryngii (PE24 and NC24 respectively). Estimate = log2 fold change; the vertical red lines are indicative and are set to fold change = 1.5); the horizontal red line is set to FDR = 0.05.