PLoS ONE (Jan 2014)
Composition of intestinal microbiota in immune-deficient mice kept in three different housing conditions.
Abstract
Abundance of commensals constituting the intestinal microbiota (IM) affects the immune system and predisposes to a variety of diseases, including intestinal infections, cancer, inflammatory and metabolic disorders. Housing conditions determine the IM and can hence influence the immune system. We analyzed how both variables affect the IM of four immune-compromized mouse lines kept under different housing conditions.We investigated the IM composition in mice by quantitative 16S rRNA RT-PCR analysis of the main fecal bacterial groups (Enterobacteriaceae, enterococci, lactobacilli, bifidobacteria, Bacteroides/Prevotella (BP) spp., Clostridium leptum and coccoides groups). Mice were homozygous (HO) or heterozygous (HE) for a targeted inactivating mutation of either the IFN-γ Receptor (R), IFN-γ, Rag1 or IL-4 genes. Overall, differences in IM composition were subtle. However, in the SPF-barrier, total eubacterial loads were higher in Rag1 HE versus Rag1 HO mice as well as in IFN-γR HE versus IFN-γR HO and WT animals. Although absent in WT mice, bifidobacterial loads were higher in HO and HE IFN-γ and Rag1 as well as IL-4 HO mice. Furthermore, BP was slightly lower in HO and HE IFN-γR and IFN-γ mice as well as in IL-4 HO mice as compared to WT controls. Interestingly, IM compositions were comparable in WT mice when kept in individual ventilated cages (IVC) or open cages (OC). IFN-γ HO and HE mice, however, had higher enterobacteria and BP loads, but lacked bifidobacteria when kept in OC versus IVC, as was the case in HO and HE Rag1 mice. In addition, Rag1 HO mice harbored higher clostridial loads when housed in OC as compared to IVC. Unexpectedly, lactobacilli levels were higher in IFN-γR mice when kept in OC versus IVC.Housing-dependent and immune-deficiency mediated changes in intestinal microbiota composition were rather subtle but may nevertheless impact immunopathology in experimental models.