Dextran sodium sulfate-induced colitis alters the proportion and composition of replicating gut bacteria

Abstract

The bacteria living in the human gut are essential for host health. Though the composition and metabolism of these bacteria are well described in both healthy hosts and those with intestinal disease, less is known about the metabolic activity of the gut bacteria prior to, and during, disease development-especially regarding gut bacterial replication. Here, we use a recently developed single-cell technique alongside existing metagenomics-based tools to identify, track, and quantify replicating gut bacteria both ex vivo and in situ in the dextran sodium sulfate (DSS) mouse model of colitis. We show that the proportion of replicating gut bacteria decreases when mice have the highest levels of inflammation and returns to baseline levels as mice begin recovering. In addition, we report significant alterations in the composition of the replicating gut bacterial community ex vivo during colitis development. On the taxa level, we observe significant changes in the abundance of taxa such as the mucus-degrading Akkermansia and the poorly described Erysipelatoclostridium genus. We further demonstrate that many taxa exhibit variable replication rates in situ during colitis, including Akkermansia muciniphila. Lastly, we show that colitis development is positively correlated with increases in the presence and abundance of bacteria in situ which are predicted to be fast replicators. This could suggest that taxa with the potential to replicate quickly may have an advantage during intestinal inflammation. These data support the need for additional research using activity-based approaches to further characterize the gut bacterial response to intestinal inflammation and its consequences for both the host and the gut microbial community.IMPORTANCEIt is well known that the bacteria living inside the gut are important for human health. Indeed, the type of bacteria that are present and their metabolism are different in healthy people versus those with intestinal disease. However, less is known about how these gut bacteria are replicating, especially as someone begins to develop intestinal disease. This is particularly important as it is thought that metabolically active gut bacteria may be more relevant to health. Here, we begin to address this gap using several complementary approaches to characterize the replicating gut bacteria in a mouse model of intestinal inflammation. We reveal which gut bacteria are replicating, and how quickly, as mice develop and recover from inflammation. This work can serve as a model for future research to identify how actively growing gut bacteria may be impacting health, or why these particular bacteria tend to thrive during intestinal inflammation.

Keywords: bacterial replication; bioinformatics; click chemistry; gut microbiome; whole-genome sequencing.

Fig 1

Experimental timeline and changes in the gut bacterial community during DSS-colitis. (A) Timeline for each mouse experiment. After a 2-week acclimatization period, the feces of four cages of mice (two cages per sex, with 3 mice per cage) were collected over 7 days. Then, DSS was administered in the drinking water of the mice at a concentration of 2% (wt/vol) for 5 days, after which normal water was returned. Mouse feces continued to be collected throughout. Once DSS administration began, and thereafter, mouse health was monitored via measurements of blood in stool and lipocalin-2. Based on mouse health parameters, the phases of this study were split into four health states: baseline, pre-symptomatic, symptomatic, and recovery. (B) Taxonomic bar plots at the genus level of the fecal gut bacteria from each mouse during each health state (C) Beta diversity (Jaccard index) of the fecal gut bacterial communities from all cages, grouped per health state. (D) Heat maps of standardized relative abundances of taxa from all cages of mice which were differentially abundant as compared to the baseline state. BA = Baseline; PS = Pre-symptomatic; SY = Symptomatic; RE = Recovery.

Fig 2

Experimental workflow and changes in the proportion and composition of replicating gut bacteria during DSS colitis. (A) Workflow for each mouse experiment. Whenever fecal pellets were collected, half were immediately used for EdU-click and FACSeq; the other half were frozen for later DNA extraction, whole-genome shotgun sequencing, and subsequent analyses. (B and C) The median proportion of replicating (%EdU+) cells, represented as a line graph showing per-cage dynamics over time (B) and as boxplots, grouping all cages per health state (C). (D and E) Beta diversity (Jaccard indices) of the replicating (EdU+) (D) and whole community (Whole) (E) of bacterial cells for all cages, grouped by health state.

Fig 3

Taxonomy and differentially abundant taxa during DSS colitis. (A and B) Taxonomic bar plots of the fecal gut bacteria for each cage during each health state for the replicating (EdU+) (A) and whole community (Whole) (B) of bacterial cells. (C and D) Heat maps of the standardized abundance of taxa were determined as differentially abundant as compared to the baseline state for the replicating (EdU+) cells (C) and the whole community (Whole) (D).

Fig 4

Relative abundance and differentially abundant taxa from whole genome shotgun sequencing (WGS) data. (A and B) Taxonomic bar plots at the genus level for the metagenome-assembled genomes. (C) Volcano plots of metagenome-assembled genomes were determined to be differentially abundant from the baseline. Note that genomes 14-2 and COE1 are both from the Lachnospiraceae family.

Fig 5

Replication dynamics of gut bacteria during DSS colitis. (A–D) Replication rates of representative taxa as calculated by GRiD, colored by health state. (E) Pearson correlation between health states and minimal doubling times (DTs) for both cages, as calculated by gRodon.