Introduction Helicobacter felisinfection of C57BL/6 mice is an established model of human gastric carcinogenesis. Mice infected with H. felisby orogastric gavage show acute gastritis, and subsequently develop pre-neoplastic pathology. The acute component of this model mimics the pathology that has been identified in cases of acute Helicobacter pylori infection in humans, however most of these reports describe the consequences of ingestion of large H. pyloriinocula. It is not clear how this reflects the natural acquisition of H. pyloriwhich is usually a subclinical event. Previous studies have reported that co-housing H. felisinfected mice with uninfected littermates did not lead to H. felistransmission, but these experiments were performed before molecular diagnostic tests for bacterial colonisation were available.

Method A C57BL/6 mouse, conventionally infected withH. felis, was co-housed with 9 uninfected littermates. Prokaryotic DNA was extracted from stool samples collected from individual mice at regular intervals. Mice were culled at week 7.Gastric mucosal samples were collected for nucleic acid and histological assessment. H. felistransmission was assessed by H+E, modified Giemsa and Warthin-Starry stains and by quantitative PCR (qPCR) for the H. felisgene flaA. Longitudinal changes in faecal microbiota were assessed by qPCR for selected bacterial phyla.

Results FlaA was identified by qPCR in gastric and stool samples from the actively infected mouse. All 3 histological stains identified H. felisin the gastric antrum of this mouse, its gastric corpus exhibited early atrophy and inflammation. qPCR of stool samples from the other 9 mice did not identify flaA DNA. In keeping with previous studies, H. felis was not detected histologically in the gastric antral mucosa of these mice and the gastric corpus mucosa wasmorphologically normal. However flaA template DNA was identified in gastric mucosal samples from 2 of the 9 mice.Selected bacterial phyla were quantified from stool samples taken longitudinally from each mouse. PCA analysis demonstrated that the colonic microbiota of mice converged over time. A particular shift in faecal microbiota was observed in the actively infected mouse. No differences were identified in the mice that had been passively colonised with H. feliscompared to uninfected mice.

Conclusion H. feliswas transmitted between co-housed mice in SPF conditions. Passive colonisation led to little inflammation and no gross shift in faecal microbiotal composition was seen. This may represent a model of asymptomatic human H. pyloriacquisition. Further investigation using larger cohorts of animals maintained for longer times will address whether the mode of acquisition of H. felis influences eventual pathological outcome.

Disclosure of interest None Declared.