| Objectives:The aims of current study are to establish different types of in vitro gut modeling systems and study the potential application to stimulate the human intestinal microbiota. Methods: ?A batch in vitro system was applied to screen the effects of different carbohydrates on the alcohol production in human fecal samples. The stools of the auto-brewery syndrome patient and two healthy volunteers were inoculated in the media containing different carbon sources, including glucose, starch or dextrin. After 24 hours fermentation, the alcohol and VFA concentrations were detected by GC and HPLC respectively. Then fecal samples were inoculated into medium containing rifaximin, metronidazole, levofloxacin and imipenem to test the effects of antibiotics on the alcohol production. ?A single-stage chemostat system with working volume 330ml was established. The system was maintained at pH 6.2 and dilution rate was set up at 0.04/h. Stool samples from 6 volunteers were collected and inoculated into the systems, in which starch medium (VI) and starch polysaccharide medium (XP) were used. The stability of bacterial communities inside the systems was checked by PCR-DGGE analysis. The similarities of bacterial communities between fecal inocula and the fermentation products were determined by sequencing 16s rRNA genes. ? Finally, the effects of fecal bacteria and its fermentation products on the inflammatory responses were tested by DSS mouse model. In brief, C57BL/6 mice were fed with 2% DSS for six days to induce the ulcers in the colon. The mortalities were recorded daily after oral gavaging fecal inocula and fermentation products. Results: ?Patient stool produced higher concentration of alcohol at 24h in ?-glucose medium than healthy controls did (P<0.05), and ?-glucose was best carbon source to sustain the alcohol production than starch and dextrin did (P<0.05). Alcohol was still produced by patient stool when inoculated in Rifaximin, Metronidazole, Levofloxacin containing medium, but was hardly detected in Imipenem.16s rRNA analysis showed that Escherichia-Shigella, Parasutterella and Veillonella were bacteria that might be responsible for the alcohol production.?In second part of study, the single stage chemostat system reached steady stage after 8 days fermentation, based on the patterns of PCR-DGGE analysis. Total 80 to 160 OTUs were detected from each sample; however, the OTU numbers were less in fermentation samples than those in stool inocula, suggesting the in vitro system can't completely mimic the fecal communities. For each fecal sample, XP media generally produced more OTUs than that in VI media. Stool sample TXZ, FB, SQS were classified as Bacteroides enterotype, since the bacterial community was mainly composed by Bacteroides, while LW, WYS, CXX had more Prevotella. Bacteroides, Firmicutes, Proteobacteria and Actinobacteria were the predominated phylua in the fecal samples, while fermentation sample was mainly composed of Bacteroides, Firmicutes, Proteobacteria and Fusobacteria. For the calculation of similarity, approximately 50%of species from fecal inocula can grow in the VI and XP medium. Moreover, the enterotypes presented in the fecal inocula also influenced the simulation similarity. When considering the influence of bacterial abundance, the similarities calculated from Bateroides enterotype inocula, (TXZ, FB and SQS) reached 80%-90%in VI and XP media, while similarities was around 25%for Prevotella enterotype (LW, CXX and WYS). The two growth media could selectively enrich certain OTUs, some of which were hardly detected in stool samples due to the low abundance. The SCFA concentration in fermenters was 10-30 mmol/L and propionate and butyrate were the major SCFA.?Finally, the effects of fecal inocula with different enterotypes and the corresponding fermentation products on the development of ulcers were evaluated using DSS mice. Higher mortalities and disease activity indexes were observed in the groups of mice intake of both Bacteroides (group C) and Prevotalla (group D) enterotypes feces. However, the mortalities and disease activity indexes were similar to the control (group B, DSS treated group) when the corresponding fermentation products, group E and F were inoculated into the mice by oral gavaging. Conclusions:Current study demonstrated that in vitro batch fermentation system can be used for quick screening the enteric pathogens for clinical samples. The in vitro human gut modeling system is good at simulation of human fecal sample with the Bacteroides enterotype under current growth condition. The similarity for simulation of the feces with Bacteroides enterotype reached 80%, while the simulation for the Prevotella enterotype is only 25%. Improvement of the growth conditions is needed in the future to increase the simulation efficiency for Prevotella enterotype feces. Moreover, the effects of feces with different enterotypes on the UC development such as mortality rates and inflammatory status also need to be investigated. |
PDF Full Text Request