Effects of Antibiotics on the Intestinal Microbiota of Mice

Abstract

Studies on human and mouse gastrointestinal microbiota have correlated the composition of the microbiota to a variety of diseases, as well as proved it vital to prevent colonization with resistant bacteria, a phenomenon known as colonization resistance. Antibiotics dramatically modify the gut community and there are examples of how antibiotic usage lead to colonization with resistant bacteria [e.g., dicloxacillin usage selecting for ESBL-producing E. coli carriage], as shown by Hertz et al. Here, we investigated the impact of five antibiotics [cefotaxime, cefuroxime, dicloxacillin, clindamycin, and ciprofloxacin] on the intestinal microbiota in mice. Five different antibiotics were each given to groups of five mice. The intestinal microbiotas were profiled by use of the IS-pro analysis; a 16S-23S rDNA interspace [IS]-region-based profiling method. For the mice receiving dicloxacillin and clindamycin, we observed dramatic shifts in dominating phyla from day 1 to day 5. Of note, diversity increased, but overall bacterial load decreased. For ciprofloxacin, cefotaxime, and cefuroxime there were few overall changes. We speculate that antibiotics with efficacy against the abundant anaerobes in the gut, particularly Bacteroidetes, can in fact be selected for resistant bacteria, disregarding the spectrum of activity.

Keywords: 16S; IS-pro; antibiotics; dicloxacillin; intestinal microbiota.

Figure 1

Bacterial phylum profiles per mouse and antibiotic group (five mice per antibiotic or control), established by 16S rRNA analyses. Colour intensity of the bands indicates quantities of respective species. Each band represents a microbial interspace (ITS) fragment, corresponding to an operational taxonomy unit (OTU). Fragments are sorted into their corresponding phylum. No bands; indicate the absence of bacterial species. “1” = Day 1, “3” = Day 3, and “5” = Day 5. Numbers below indicate that data are shown for each individual mouse. Five mice in two different cages. Blue boxes indicate the shift in Bacteroidetes composition for clindamycin and dicloxacillin. Red boxes indicate the shift in Firmicutes composition for ciprofloxacin, clindamycin, and dicloxacillin. Black boxes show depletion of Proteobacteria in most antibiotics groups (and in controls). We have indicated the concurrent invasion of E. coli in clindamycin and dicloxacillin treated mice. Control: No antibiotics given; CFR: Cefuroxime; CIP: Ciprofloxacin; CLI: Clindamycin; CTX: Cefotaxime; DIC: Dicloxacillin.

Figure 2

Box plot of the similarity of gut microbiota at day 3 and 5 compared to the initial microbiota. The y-axis represents cosine similarity values, for pairs of IS-profiles of the microbiota at different dates (higher value means higher similarity to pretreatment microbiota). Shown here is the normal variation and the variation induced by antibiotic treatment. Boxes represent the cosine similarity values for all five mice in each group.

Figure 3

The principal coordinates analysis (PCoA) based on cosine distance between samples from all mice on all days. Each dot represents a gut microbiota profile. Colours show the antibiotic group and day. The further dots are separated, the larger the dissimilarity between associated gut microbiota profiles. The group to the right (red and orange) represent individual profiles from mice in the clindamycin and dicloxacillin groups at days 3 and 5. These are clearly separated from the rest of the samples. PCoA starts by putting the first point at the origin, and the second along the first axis (e.g., PC1). Each axis can represent a phylum or bacterial species. Further points can be added, which usually means adding a second axis followed by a possible third axis and so forth (e.g., PC2 and PC3 etc.). A successful PCoA will, therefore, generate a 2–3 axes. Each object has a ‘score’ along each axis. The combined scores delivering the position in the plot. Via PCoA, we visualize the dissimilarities, or distances, between samples from mice receiving antibiotics. Interpretation of a PCoA plot is straightforward: Objects closer to one another are more similar than those further away. The points illustrate the normal variation for controls and the variation induced by antibiotic treatment. Antibiotic treatments were given once a day for three days (day 1 to day 3). The faeces was collected on day 1 (prior to the antibiotic treatment), on day 3, and finally on day 5 (two days after the end of treatment), respectively. “1” = Day 1, “2” = Day 2, and “5” = Day 5. Control: No antibiotics given; CFR: Cefuroxime; CIP: Ciprofloxacin; CLI: Clindamycin; CTX: Cefotaxime; DIC: Dicloxacillin.