Colon simulator model

For modelling the infant gut microbiota, frozen infant faecal samples were used to prepare an inoculum of the colon simulator system. Donor infants, all aged under 1 year, were in good health and had not been medicated with antibiotics. A parent of each infant gave informed consent and provided background information on the infant who was donating the faecal sample (age, food, supplements, allergies, and delivery mode). Parents were provided with the instructions and equipment for sample handling. Infant faecal samples were frozen immediately at home before the samples were collected from each family. The study was reviewed and approved by the Coordinating Ethical Committee of the University of Helsinki (Decision number 139/13/03/00/16). All methods were in accordance with the national guidelines in Finland.

To study the effect of 2′-FL on the infant intestinal microbiota, the four-stage semi-continuous EnteroMIX® colon simulator model was used^14,16,17. The technical specifications of this in vitro model have been described earlier^16,67. In brief, the simulator consists of four units, in which four simulations from the same inoculum can be run simultaneously and in parallel. A single unit of the EnteroMIX® colon simulator contained four sequentially connected glass vessels, V1 to V4, representing the different compartments of the human colon, ranging from the ascending colon to the sigmoid/rectum area. The volume of microbial slurry increased from V1 (6 ml) to V4 (12 ml) to mimic reduced flow. pH levels (pH 5.5, 6.0, 6.5 and 7.0 for vessels V1, V2, V3, and V4, respectively) were controlled and adjusted with 1% ammonia. The entire system was maintained at 37 °C in an anaerobic atmosphere. All run parameters were controlled with a computer using customised software. Artificial ileal fluid was used as a medium in the EnteroMIX® colon simulator⁶⁷. 2′-FL (DuPont Nutrition and Health, Kantvik, Finland and Inbiose, Ghent, Belgium), lactose (Sigma-Aldrich, St. Louis, MO, USA) and GOS (Clasado Biosciences, St Helier, Jersey, United Kingdom) were suspended in artificial ileal fluid (2% concentration, w/v) to serve as the sole carbon source. Artificial ileal fluid alone was fed to the system for control simulations.

Parents froze the collected faecal samples at −20 °C, and the samples were stored at −80 °C in the laboratory until used as inoculum. When required to generate enough inoculum for the four units of the simulator, faecal samples from the same donor collected within one week were pooled. The faecal samples from a single donor were mixed with 3 parts (wt/wt) of artificial ileal fluid, filtered through 0.3 mm metal mesh and incubated anaerobically at +37 °C for 24 hours, and then, this faecal slurry was added to the simulator units. Samples from the slurry were taken to determine the composition of the inoculum. The test products (2′-FL, GOS, lactose) or control were fed into the simulator system at 3-hour intervals during the simulation for a total time of 48 hours. Samples were then collected from the simulator vessels; the composition of the simulated microbiota, and microbial metabolites were analysed.