BACKGROUND: To properly assess the robustness and translatability of a bacterial live biotherapeutic product (LBP) to survive and perform in the gastrointestinal tract, a complex microbiota is required. Direct bacteria-to-bacteria interactions, competition, and immunological states, all which can mediate the dynamics between an LBP and the host, are not recapitulated in a germ-free or reduced gut community. We aim to specifically evaluate competition between an introduced LBP and its parental strain; the principle of competitive exclusion indicates that whichever strain establishes in the niche first will take preference and exclude the secondary strain from engrafting. We hypothesize that host diet, introduction of transgenes into the LBP, and presence of other non-familial microbes will affect the competition dynamics between the LBP and its parental strain.

METHODS: Native Escherichia coli was isolated from the stool of a conventionally raised (CR) C57BL/6 mouse and engineered to express green fluorescent protein (GFP) and kanamycin resistance (EcAZ-2). Our published work demonstrates that native E. coli are superior to lab-strain probiotics in functionally manipulating the gut microbiome. EcAZ-2 was engineered to either express spectinomycin resistance, bile salt hydrolase (BSH) with chloramphenicol resistance, or interleukin-10 (IL10) with chloramphenicol resistance. In vitro growth rates were measured to confirm no major fitness deficits were produced by engineering the bacteria. In vivo competition studies were conducted by administering two strains (EcAZ-2^SpecRand either EcAZ-2^BSHor EcAZ-2^IL10) to CR C57BL/6 mice via singular oral gavages, varying the timing and order of delivery. Impact of different diets such as normal chow, high fat diet, or atherogenic diet was also examined.

RESULTS: EcAZ-2 colonizes the entirety of the intestines for the duration of a mouse’s lifetime. The addition of a transgene does not affect engraftment and transgene function is maintained for the length of engraftment.
Order of strain introduction was not sufficient to determine competition outcomes nor exclude the second strain from engrafting in CR mice, but it could affect the final proportion of the strains. Furthermore, luminal conditions (altered by diet composition, inflammation, etc) affected these competition dynamics and outcomes.
Surprisingly, colonization of a CR mouse by EcAZ-2 can be prevented if a mouse is already colonized with a particular Enterococcus strain (EcrcAZ).

CONCLUSIONS: Our results directly contradict the colonization exclusion paradigm that was established in gnotobiotic mice, indicating that bacterial competition findings in gnotobiotic mice cannot be extrapolated to CR hosts. Identifying factors that affect bacterial fitness and competition within the gut will be important in the development of efficacious and durable LBPs.

Nicole Siguenza

Appearances