Role of the microbiota in the defense against infections by Enterococci
Antibiotic resistant bacteria, such as vancomycin-resistant Enterococcus (VRE) are an increasing problem in hospitalized patients and commonly cause infections following antibiotic therapy. Infections with VRE generally begin by colonization of the intestinal tract. In normal conditions, our gut is colonized by hundreds of commensal bacterial species, the microbiota, that suppress intestinal colonization by VRE, a phenomenon known as “colonization resistance” (CR). However, administration of antibiotics alters the composition of the microbiota, which allows VRE to densely colonize the intestine and subsequently disseminate to the bloodstream where it can put in serious danger the life of the patient. Thus, understanding how and which members of the microbiota confer CR and how antibiotics promote intestinal colonization by VRE, is crucial if we want to prevent VRE infections. Unfortunately, the absence of techniques to study complex bacterial populations has hampered, until the last recent years, the research in this clinically relevant field. For this reason, it is not completely understood how antibiotics change the microbiota and promote infections by VRE, which are the members of the microbiota that are key for conferring colonization resistance against VRE and the mechanisms by which they confer protection. Thus in this thesis, the main objectives proposed are (i) to understand how antibiotics change the composition of the microbiota and subsequently promote intestinal colonization by VRE, (ii) identify commensal bacterial species that are key for conferring protection and (iii) study mechanisms by which these commensal bacteria may confer protection against VRE. We first studied the effect of oral vancomycin treatment on the human gut microbiota through 16s rRNA high-throughput sequencing. This antibiotic is frequently given to patients to treat Clostridium difficile infections and subsequently can promote secondary infections by VRE. Our analysis showed that vancomycin promotes drastic changes on the composition of the microbiota, including the depletion of all bacterial species from the phylum Bacteroidetes, one of the most prevalent phyla inhabiting in the human intestinal tract. Moreover, our analysis indicates that the microbiota never recovers its baseline state, even after 22 weeks post-antibiotic cessation. Importantly, the microbiota recovery rate was different depending on the subject analyzed. While some patients recovered most of their baseline bacterial species, up to 89% of their baseline bacterial species were not recovered in other patients. Moreover, using a mouse model we were able to demonstrate that the microbiota recovery rate upon vancomycin treatment was clinically relevant since a lower microbiota recovery upon vancomycin cessation is associated with an increased susceptibility to VRE intestinal colonization. Subsequently, using the same VRE infection mouse model, we analyzed the impact of other antibiotics of different spectrum (ciprofloxacin, neomycin, ceftriaxone, ampicillin, clindamycin, besides vancomycin) on the gut microbiota composition and on the VRE colonization capacity, both during the treatment and two weeks after antibiotic cessation. Our analysis shown, as expected, that different antibiotics promoted different changes in the composition of the microbiota, being vancomycin and clindamycin those that promoted a higher number of changes, while neomycin or ciprofloxacin had a minor effect on the microbiota composition. The changes in the composition of the microbiota were associated with the capacity of VRE to colonize the intestinal tract since those antibiotics that promote more alterations on the microbiota allowed a higher level of VRE intestinal colonization. Subsequently, applying correlation analysis and Linear Discriminatory analysis, we were able to identify several bacterial taxa that are associated with VRE resistance. These taxa include the genera Alistipes, Barnesiella, Oscillibacter and some members of the families Lachnospiraceae and Ruminococcaceae. We next isolated and administered these bacteria to vancomycin-treated mice to test their capability to restore colonization resistance. We demonstrated that the administration of a combination of 4 bacterial isolates to mice (Alistipes, Barnesiella, Oscillibacter and a bacterium from Ruminococcaceae family) drastically diminished the capacity of VRE to colonize the intestinal tract of antibiotic-treated mice. We next performed metatranscriptomic and metabolomic analysis to identify in vivo functions expressed and metabolites produced by the identified protective commensal bacteria in order to determine possible mechanisms by which these bacteria could be suppressing VRE intestinal colonization (e.g. production of inhibitory molecules, competition for nutrients). In addition, we also analyzed the in vivo transcriptome of VRE to determine the functions express by this pathogen to colonize the intestinal tract, and we performed nutrient arrays to identify nutrients that the pathogen could be using for its growth in the intestinal tract. The analysis performed determined that (i) oral inoculation of mice with the protective bacterial mixture restores the expression of genes encoding for transporters that internalize saccharides (cellobiose, N-acetyl-galactosamine, N-acetyl-glucosamine) and amino acids (serine), whose expression was diminished upon antibiotic treatment. (ii) Increased expression of these genes was associated with diminished intestinal availability of several nutrients including saccharides (i.e. cellobiose) and amino acids (i.e. serine, proline, leucine and threonine). (iii) Some of the nutrients consumed upon administration of the protective commensal bacteria may be crucial for VRE growth in the intestinal tract. Indeed, analysis of the VRE transcriptome in colonized mice revealed that VRE highly express in vivo transporters for the internalization of saccharides (i.e. cellobiose, N-acetyl-galactosamine, N-acetyl-glucosamine) and several branched-chain aminoacids (i.e. proline). Moreover, out of 190 carbon sources tested, the saccharides cellobiose, N-acetyl-glucosamine and N-acetyl-galactosamine are among the 10 carbon sources that promote the highest VRE growth under anaerobic conditions. Thus, although additional experiments should be performed for validation, our results support a model in which the administered protective bacteria decrease the intestinal levels of nutrients that can be utilized by VRE for growth, which reduces the capacity of this pathogen to colonize the intestinal tract.