The recent demonstration that DNA from several mucosal bacteria, including Prevotella spp, could be found in numerous tissues (mesenteric lymph nodes, spleen, serum, liver, lung, eye and ankle joints), either in HLA-B27 rats with or without arthritis, or control rats without HLA-B27, could be a revolution in our understanding of spondyloarthritis and close disorders, including rheumatoid arthritis. Indeed, similar translocations of dead or alive bacteria or fungi from mucosa to joints, could contribute to the onset and flares of inflammatory rheumatisms. This state of the art article addresses six questions revived by this finding: 1-How does this bacterial DNA or living bacteria traffic from mucosa to joints? 2-Can some mucosal bacteria remain alive in those tissues, including joints? 3-Could bacteria from the gut microbiota (‘self-bacteria’) protect the host cells from invasion by more pathogenic bacteria (like dog-shepherds protect from wolves)? 4-Does the composition of the joint or bone marrow microbiota depends on local metabolism, which might differ from gut metabolism? 5-Could bacterial antigens from mucosal microbiota be sufficient to trigger trained immunity of presenting cells in joints, or does such phenomenon (with lasting epigenetic changes of presenting cells) require intra-cellular infection of presenting cells or their ancestors? 6-In which subsets of cells could living bacteria preferentially persist for a long period in the joint area? Transient or dormant infections within bone-marrow mesenchymal stem cells leading to trained immunity of some of their daughter cells in joints or enthesis, lasting after clearance or the invader, is an attractive hypothesis to test.