LAUSR

Escherichia coli and Enterococcus faecalis have been implicated as important players in human gut health that have been associated with the onset of inflammatory bowel disease (IBD). Bacteriophage (phage) therapy has been used for decades to target pathogens as an alternative to antibiotics, but the ability of phage to shape complex bacterial consortia in the lower gastrointestinal tract is not clearly understood. We administered a cocktail of six phages (either viable or heat-inactivated) targeting pro-inflammatory Escherichia coli LF82 and Enterococcus faecalis OG1RF as members of a defined community in both a continuous fermenter and a murine colitis model. The two target strains were members of a six species simplified human microbiome consortium (SIHUMI-6). In a 72-h continuous fermentation, the phage cocktail caused a 1.1 and 1.5 log (log10 genome copies/mL) reduction in E. faecalis and E. coli numbers, respectively. This interaction was accompanied by changes in the numbers of other SIHUMI-6 members, with an increase of Lactiplantibacillus plantarum (1.7 log) and Faecalibacterium prausnitzii (1.8 log). However, in germ-free mice colonized by the same bacterial consortium, the same phage cocktail administered twice a week over nine weeks did not cause a significant reduction of the target strains. Mice treated with active or inactive phage had similar levels of pro-inflammatory cytokines (IFN-y/IL12p40) in unstimulated colorectal colonic strip cultures. However, histology scores of the murine lower GIT (cecum and distal colon) were lower in the viable phage-treated mice, suggesting that the phage cocktail did influence the functionality of the SIHUMI-6 consortium. For this study, we conclude that the observed potential of phages to reduce host populations in in vitro models did not translate to a similar outcome in an in vivo setting, with this effect likely brought about by the reduction of phage numbers during transit of the mouse GIT.