Mechanistic Insights Into the Cross-Feeding of Ruminococcus gnavus and Ruminococcus bromii on Host and Dietary Carbohydrates

Abstract

Dietary and host glycans shape the composition of the human gut microbiota with keystone carbohydrate-degrading species playing a critical role in maintaining the structure and function of gut microbial communities. Here, we focused on two major human gut symbionts, the mucin-degrader Ruminococcus gnavus ATCC 29149, and R. bromii L2-63, a keystone species for the degradation of resistant starch (RS) in human colon. Using anaerobic individual and co-cultures of R. bromii and R. gnavus grown on mucin or starch as sole carbon source, we showed that starch degradation by R. bromii supported the growth of R. gnavus whereas R. bromii did not benefit from mucin degradation by R. gnavus. Further we analyzed the growth (quantitative PCR), metabolite production (¹H NMR analysis), and bacterial transcriptional response (RNA-Seq) of R. bromii cultured with RS or soluble starch (SS) in the presence or absence of R. gnavus. In co-culture fermentations on starch, ¹H NMR analysis showed that R. gnavus benefits from transient glucose and malto-oligosaccharides released by R. bromii upon starch degradation, producing acetate, formate, and lactate as main fermentation end-products. Differential expression analysis (DESeq 2) on starch (SS and RS) showed that the presence of R. bromii induced changes in R. gnavus transcriptional response of genes encoding several maltose transporters and enzymes involved in its metabolism such as maltose phosphorylase, in line with the ability of R. gnavus to utilize R. bromii starch degradation products. In the RS co-culture, R. bromii showed a significant increase in the induction of tryptophan (Trp) biosynthesis genes and a decrease of vitamin B12 (VitB12)-dependent methionine biosynthesis as compared to the mono-culture, suggesting that Trp and VitB12 availability become limited in the presence of R. gnavus. Together this study showed a direct competition between R. bromii and R. gnavus on RS, suggesting that in vivo, the R. gnavus population inhabiting the mucus niche may be modulated by the supply of non-digestible carbohydrates reaching the colon such as RS.

Keywords: Ruminococcus; cross-feeding; gut bacteria; mucin; resistant starch.

FIGURE 1

Growth curves of the mono- and co-cultures with mucin (A), soluble starch (SS) (B) or resistant starch (RS) (C) as sole carbon source and cell concentrations in the different growth conditions (D). The concentrations were determined by qPCR and expressed as 16S rDNA copy number/mL of culture. The values are averages of 3 replicates for R. gnavus ATCC 29149 grown with Glc or 4 replicates for the other conditions. The error bars correspond to the standard deviations. Cells samples were collected at a time of growth of 7 h for R. gnavus ATCC 29149 grown on Glc, 10 h for R. bromii L2-63 grown on SS and 8 h for the other conditions.

FIGURE 2

Concentration of starch degradation products in the spent media. The concentrations were determined by ¹H NMR and the values are averages of 2 to 5 replicates. The error bars correspond to standard deviations. Results presented in the blue, red and green boxes come from growth assays with Glc, SS and RS as sole carbon source, respectively. Results from the YCFA medium alone, without carbon source, are presented inside the yellow box. Abbreviations: exp, exponential; sta, stationary; n/a, non-applicable.

FIGURE 3

Concentration of different metabolites in the spent media. Concentrations of ethanol, formate and acetate are shown in panel (A) while concentrations of propane-1, 2-diol and propanol are shown in panel (B). These concentrations were determined by ¹H NMR and the values are averages of 2 to 5 replicates. The error bars correspond to standard deviations. Results presented in the blue, red and green boxes correspond to growth assays with Glc, SS and RS as sole carbon source, respectively. Results from the YCFA medium alone, without carbon source, are presented inside the yellow box. Abbreviations: exp, exponential; sta, stationary; n/a, non-applicable.

FIGURE 4

Principal component analysis (PCA) plots for transcriptomics data of R. bromii L2-63 genes (A) and R. gnavus ATCC 29149 genes (B).

FIGURE 5

Volcano plots representing the differential expression analysis of R. bromii L2-63 genes. Genes are considered to be differentially expressed when Log2 Fold Change < –1.5 or > 1.5 and padj < 0.05; non-differentially expressed genes are shown as blue dots. The impact of starch on R. bromii L2-63 gene transcription in mono-cultures and co-cultures is shown in panel (A,B), respectively. (A) No gene was differentially expressed between both mono-cultures. (B) When comparing the co-cultures, 11 genes were up-regulated with RS as compared to SS (shown as green dots). The impact of R. gnavus ATCC 29149 on R. bromii L2-63 gene transcription with SS and RS as sole carbon source is shown in panel (C,D), respectively. (C) When SS was used as carbon source, 7 genes were up-regulated in the co-culture as compared to the mono-culture (shown as green dots). (D) When RS was used as carbon source, 23 genes were up-regulated in the co-culture (shown as green dots) while 4 genes were upregulated in the mono-culture (shown as red dots).

FIGURE 6

Heatmap of the transcription level (in arbitrary unit) of differentially expressed (Log2 Fold Change < –1.5 or > 1.5 and padj < 0.05) R. bromii L2-63 genes in different growth conditions. This heatmap was produced with ClustVis web tool (Metsalu and Vilo, 2015) using the transcript counts as input values.

FIGURE 7

Volcano plots representing the differential expression analysis of R. gnavus ATCC 29149 genes. Genes were considered to be differentially expressed when Log2 Fold Change < –1.5 or > 1.5 and padj < 0.05; non-differentially expressed genes are shown as blue dots. Panel (A) shows the impact of starch type on R. gnavus ATCC 29149 gene transcription when co-cultured with R. bromii L2-63; 213 genes were upregulated in the co-culture with RS (shown as green dots) while 212 genes were up-regulated in the co-culture with SS (shown as red dots). The combined effect of the presence of R. bromii L2-63 and the carbon source (starch vs. glucose) is shown in panels (B) and (C) when SS or RS was used in the co-culture, respectively; (B) When SS was used as carbon source, 40 genes were up-regulated in the co-culture (shown as green dots) and 59 genes were up-regulated in the mono-culture (shown as red dots). (C) When RS was used as carbon source, 119 genes were up-regulated in the co-culture (shown as green dots) while 101 genes were up-regulated in the mono-culture (shown as red dots).

FIGURE 8

Heatmap of the transcription level (in arbitrary unit) of selected differentially expressed (Log2 Fold Change < –1.5 or > 1.5 and padj < 0.05) R. gnavus ATCC 29149 genes in different growth conditions. This heatmap was produced with ClustVis web tool (Metsalu and Vilo, 2015) using the transcript counts as input values. The 20 R. gnavus ATCC 29149 genes with an upregulated transcription in both co-cultures with R. bromii L2-63 on starch compared to the mono-culture on Glc are in blue. The 22 R. gnavus ATCC 29149 genes with an upregulated transcription in the mono-culture on Glc compared to both co-cultures with R. bromii L2-63 on starch are in black.