Resilience of Gut Microbial Community Structure
How does colonization resistance become compromised in the first place? Evidence suggests that insults to the microbiome, such as antibiotic treatment, can have long-lasting effects on the indigenous microbial community. Young and colleagues treated mice with antibiotics, let the animals recover from the antibiotic stress in a sterile environment, and then observed what happened when they were either left alone in a sterile environment or co-housed with a donor mouse (Antonopoulos et al., 2009). They found that mice left alone, with no donor mouse present to repopulate their guts, had microbiota that looked very similar to each other but very different from microbiota in mice that had been co-housed with donors. Even 6 weeks after stopping antibiotic treatment, mice left alone had much lower microbiota diversity than the other mice. However, as with humans, if fecal transplantation is done, diversity can return to normal.
Subsequent mouse research showed that with respect to C. difficile infection, colonization resistance can be overcome by the administration of specific (but not all) antibiotics or combinations of antibiotics. Although early work with antibiotic-treated mice was unsuccessful in modeling human C. difficile infection, Chen and colleagues (2008) were able to establish disease by pretreating the mice with a cocktail of five antibiotics before treating them with clindamycin and challenging them with C. difficile. Young’s team recreated the Chen et al. (2008) model and found that a pretreatment of five antibiotics without clindamycin did not cause disease, that clindamycin alone without pretreatment allowed transient colonization without disease (i.e., the infected mice shed bacteria briefly but showed no signs of inflammation), and that the combination of the pretreatment antibiotic cocktail followed by clindamycin allowed C. difficile colonization and the development of disease (Reeves et al., 2011). The severity of disease in the cocktail-plus-clindamycin treatment group varied. About half of the animals became very ill clinically, while the other half were able to maintain their health even though their gut epithelia became inflamed. The sicker animals also had more C. difficile and bacterial toxin present in their intestinal tissue.
With respect to microbial taxonomic composition, researchers observed high levels of the Firmicutes families, especially Lachnospiraceae genera (i.e., important short-chain fatty acid producers), and some Bacteroidetes families in untreated mice, but mice in the cocktail-plus-clindamycin treatment group bloomed Proteobacteria (e.g., E. coli, which Young described as only a “minority player” in a healthy gut). The microbiota in mice that developed clinical illness remained dominated by Proteobacteria over time, while the microbiota of mice that suffered some inflammation but did not become clinically ill eventually reverted to “healthy” Lachnospiraceaedominated communities (Reeves et al., 2011, 2012).