Human-derived microRNA 21 regulates indole and L-tryptophan biosynthesis transcripts in the gut commensal Bacteroides thetaiotaomicron

Abstract

In the gut, microRNAs (miRNAs) produced by intestinal epithelial cells are secreted into the lumen and can shape the composition and function of the gut microbiome. Crosstalk between gut microbes and the host plays a key role in irritable bowel syndrome (IBS) and inflammatory bowel diseases, yet little is known about how the miRNA-gut microbiome axis contributes to the pathogenesis of these conditions. Here, we investigate the ability of miR-21, a miRNA that we found decreased in fecal samples from IBS patients, to associate with and regulate gut microbiome function. When incubated with the human fecal microbiota, miR-21 revealed a rapid internalization or binding to microbial cells, which varied in extent across different donor samples. Fluorescence-activated cell sorting and sequencing of microbial cells incubated with fluorescently labeled miR-21 identified organisms belonging to the genera Bacteroides, Limosilactobacillus, Ruminococcus, or Coprococcus, which predominantly interacted with miR-21. Surprisingly, these and other genera also interacted with a miRNA scramble control, suggesting that physical interaction and/or uptake of these miRNAs by gut microbiota is not sequence-dependent. Nevertheless, transcriptomic analysis of the gut commensal Bacteroides thetaiotaomicron revealed a miRNA sequence-specific effect on bacterial transcript levels. Supplementation of miR-21, but not of small RNA controls, resulted in significantly altered levels of many cellular transcripts and increased transcription of a biosynthetic operon for indole and L-tryptophan, metabolites known to regulate host inflammation and colonic motility. Our study identifies a novel putative miR-21-dependent pathway of regulation of intestinal function through the gut microbiome with implications for gastrointestinal conditions.

Importance: The mammalian gut represents one of the largest and most dynamic host-microbe interfaces. Host-derived microRNAs (miRNAs), released from the gut epithelium into the lumen, have emerged as important contributors to host-microbe crosstalk. Levels of several miRNAs are altered in the stool of patients with irritable bowel syndrome or inflammatory bowel disease. Understanding how miRNAs interact with and shape gut microbiota function is crucial as it may enable the development of new targeted treatments for intestinal diseases. This study provides evidence that the miRNA miR-21 can rapidly associate with diverse microbial cells form the gut and increase levels of transcripts involved in tryptophan synthesis in a ubiquitous gut microbe. Tryptophan catabolites regulate key functions, such as gut immune response or permeability. Therefore, this mechanism represents an unexpected host-microbe interaction and suggests that host-derived miR-21 may help regulate gut function via the gut microbiota.

Keywords: gut microbiota; host-microbe interactions; irritable bowel syndrome; microRNAs; transcriptomics.

Fig 1

Quantification of miRNAs in fecal samples and dynamics of association of miRNAs with fecal microbiota. (A) Quantification of miR-21 and miR-26b in stool samples from colonoscopy controls (control, n = 5) and IBS-M patients (IBS, n = 6) by qPCR. All samples were normalzsed to the input amount of RNA. P-values were determined using an unpaired two-sample Wilcoxon test. (B) Schematic representation of miR-21 and miR-21 scramble control (miR-21^scr) incubations with live or ethanol-fixed (dead control) microbiota derived from fecal samples. Each miR was added to a final concentration of 250 nM. Incubations were set in duplicate and sampled immediately after amendment (0 h of incubation) or after 1, 2, 4, and 6 h of incubation. Collected samples were centrifuged and further split into a supernatant (cell-free) and a cell pellet fraction (see Materials and Methods) and processed for miRNA quantification by qPCR. (C, D) Levels of miRNA, expressed as a percentage of initial miRNA, over time in live (C) or dead control (D) microbiota incubations amended with miR-21, determined by qPCR. Three independent sets of incubations were established using fecal samples from three healthy donors (D1, D2, and D3). (E, F) Levels of miRNA, expressed as a percentage of initial miRNA, over time in live (E) or dead control (F) incubations amended with miR-21^scr, for the same three donors. Data points represent the mean of two replicates per condition, and bars represent the standard deviation. To help visualize the data displayed in the main plots, zoomed plots are displayed in panels (D, F).

Fig 2

Imaging, sorting, and sequencing of microbial populations associated or not with fluorescently labeled miRNAs. (A) Fluorescence microscopy imaging of gut microbiota cells labeled with the nucleic acid dye DAPI (shown in blue) after 1 h of incubation with miR-21- or miR-21^scr-ATTO 488 (in green). Representative images of fixed microbiota cells (dead control) incubated with miR-21-ATTO 488 are shown in the middle panel. Numbers in the top left corner indicate the percentage of cells displaying green fluorescence, and n indicates the total number of cells analyzed. Scale bar: 10 µm. (B) Family-level relative abundance profile of communities present in ATTO 488⁺ or ATTO 488^- sorted fractions, obtained after incubation with miR-21- or miR-21^scr-ATTO 488. Incubations were established from a mix of fecal samples originating from three donors, whose individual fecal microbiome composition is shown in the “donor samples” panel. Each bar represents the mean from two replicates. (C) Dendrogram summarizing the hierarchical clustering of microbial communities present in ATTO 488 + or ATTO 488- sorted fractions, as determined by 16S rRNA gene amplicon sequencing. Microbial communities present in fractions sorted based on the ATTO 488 positive (ATTO 488⁺) gate are highlighted in green, while the ones based on the ATTO 488 negative gate (ATTO 488^-) are highlighted in blue. Communities of the three individual donors from which a combined fecal slurry community was obtained are also shown as a reference. Data from two replicates per miR and per gate are shown: R1 and R2. Gr: group. (D) Alpha diversity metrics (Observed ASVs, Shannon index, and Inverse Simpson’s diversity index) in gut microbial communities described in (B) and (C). Boxes represent the median, first, and third quartiles. Whiskers extend to the highest and lowest values that are within one and a half times the interquartile range. ***P < 0.001, ****P < 0.0001; paired Wilcoxon test.

Fig 3

Microbial taxa associating with fluorescently labeled miR-21 and miR-21^scr. (A) DESeq2-based differential abundance plot showing all ASVs whose relative abundance significantly differs across both miR-21 and miR-21^scr ATTO 488⁺ gated sorts versus miR-21 and miR-21^scr ATTO 488^- gated sorts. Each data point represents an ASV grouped by genus and colored by family. (B) DESeq2-based differential abundance analysis of bacterial populations of sorted fractions for either miR-21 or miR-21^scr. All ASVs whose abundance is significantly higher in miR-21-ATTO 488⁺ compared with miR-21-ATTO 488^- sorts (filled circle), or in miR-21^scr-ATTO 488⁺ sorts compared with miR-21^scr-ATTO 488^- (open circle). In (A) and (B), the size of each data point represents the base mean counts for that ASV across the data set (DESeq2 analyses). The x-axis represents the log2 fold change in abundance of the respective ASV across treatments. Only ASVs with an adjusted P-value < 0.05 are shown.

Fig 4

Transcriptomic analyses of B. thetaiotaomicron cells supplemented with miR-21 or controls. (A) Growth of B. thetaiotaomicron in the presence of 250 nM of miR-21, miR-21^scr, a small RNA control, and no amendment (water) control. Data from three independent growths per condition are shown. Dots depict the mean per condition, and error bars represent the standard deviation. (B) Principal coordinate analysis (PCoA) based on Bray–Curtis distances summarizing transcript level profiles (read counts) of B. thetaiotaomicron cells collected 1 h after amendment with each indicated small RNA (or water as a control). (C-E) Volcano plots representing differentially abundant B. thetaiotaomicron transcripts incubated with miR-21 (C), miR-21^scr (D), and small RNA control (E) compared with the water control. In all volcano plots, the x-axis represents the log2 fold change, and the y-axis represents –log10 (adjusted P-value). All transcripts were identified and analyzed using the DESeq2, and transcripts with an adjusted P < 0.05 are represented (gray). Transcripts that show both a log2 fold change >1 or <−1 and have P < 0.05 are shown in red.

Fig 5

miR-21 differentially impacts B. thetaiotaomicron transcription and regulates transcript levels of a gene cluster involved in indole and L-tryptophan synthesis. (A) Heatmap showing significantly differentially expressed genes (DEGs) (adjusted P < 0.05) between each indicated small-RNA amended sample (relative to the water control). Only genes with a log2 fold change >1 or <−1 are shown. (B) Heatmap summarizing the log2 fold change in transcript levels in cells supplemented with miR-21 compared with all other treatments. (C) Representation of the enzymatic steps involved in indole and L-tryptophan synthesis from chorismate, highlighting intermediates and genes encoding enzymes catabolizing each step in Bacteroides spp. (44) . Colur boxes indicate the fold change in mRNA levels relative to the water controls, as determined by RT-qPCR.