Validation of gavage sampling as tool for longitudinal sampling of microbiota of the mouse gastric lumen

Fecal samples are commonly used for longitudinal studies of the gut lumen microbiome to track the course of response to infection or drug treatment, but no comparable method has been evaluated for longitudinal analysis of the gastric lumen microbiome in mice. Herein, a buffer flush of the stomach with a flexible gavage needle was used to collect gastric contents at one or several time points without harming the mouse. These samples were compared to samples collected by sacrifice and dissection of the mouse stomach. Microbiota from these samples were sequenced and evaluated in two ways: the composition of samples as measured by beta diversity and the richness of samples as measured by alpha diversity. Additionally, the effect of multiple sampling every two days on these metrics were studied. DNA was extracted from each of these samples and Illumina 16S rRNA gene sequencing was performed.

First, taxonomic richness of gavage and dissection samples was compared. A greater number of taxa was detected in gavage samples than in dissection samples. Second, taxonomic richness was analyzed over time. No significant difference in taxonomic richness was observed with repeated gavage flushes. Third, a comparison was made of the taxonomic composition of samples collected by gavage versus dissection followed by a comparison of samples collected over multiple samplings. Nonmetric multidimensional scaling analysis revealed no clear differences between collection by gavage flushing or dissection. Using weighted Unifrac and Aitchison taxonomic distances between gavage and dissection samples were not significantly different from distances between gavage samples themselves, and no significant difference was found in the taxonomic composition of mice which were sampled repeatedly. Finally, relative abundances of specific identified taxa were compared, and eleven taxa were found to differ in frequency between collection methods. Using the more stringent Analysis of Composition of Microbiomes (ANCOM), seven was found to differ. Similarly, no significant differences were uncovered using these analyses over multiple samples by gastric flush.

In summary, the consistency of the microbiota collected by gastric flushing recommends its use for microbiome analysis of gastric fluid similar to the use of fecal sampling to study the gut lumen microbiome.