Female C57BL/6 Rag-1-deficient mice (6 to 8 weeks of age) were purchased from the Charles River Breeding Laboratories (Japan). Mice were kept under specific pathogen free (SPF) conditions in animal care facility in POSTECH (Pohang, Republic of Korea). The procedure of animal experiment was approved by the Institutional Animal Care and Use Committee (IACUC) at the POSTECH Biotech Center.

Feces from wild type mice or colonic feces from H. muridarum mono-associated mice were diluted to 12 g/ml of PBS. Suspensions were placed for 3 min at room temperature. 400 µl of supernatant was injected to each Rag-1-deficient mice 2 times per week for 4 weeks by oral gavage.

Bone marrow was collected by flushing femurs with RPIM containing 2% FBS. Cells were stained with a combination of anti-lineage (CD3e, B220, Gr-1, CD11b, NK1.1, and MHC II) APC, anti-c-Kit PE-cy7, anti-CD150 PerCP-eFluor710, anti-CD48 PE (eBioscience, USA), and anti-Sca-1 BV421 (Biolegend, USA). Flow cytometry was performed on a FACS Fortessa (BD Bioscience, USA) and Canto II and the data were analyzed using Flow Jo 3.0 (Tree Star, Ashland, OR).

Each 2.5 g fecal samples were placed in FastDNA SPIN Kit lysing matrix E tubes. 978 ml of sodium phosphate buffer and 122 ml MT buffer were added to each tube and mixed. Tubes were centrifuged at 14,000 g, 10 min. Supernatant was transferred to new tubes and added 250 µl Protein Precipitation Solution (PPS) and shaken 10 times by hand. Tubes were centrifuged at 14,000 g, 5 min. Supernatant was transferred to new tubes and added 1,000 µl Binding Matrix Suspension buffer. Tubes were placed on rotator for 2 min. Tubes were placed on a rack for 3 min. 500 µl of supernatant were removed Remaining amount of supernatant were suspended and transferred to a SPIN FILTER and centrifuged at 14,000 g, 5 min. The catch tubes were emptied. 500 µl of SEWS-M were added to each tube and resuspended gently. Tubes were centrifuged at 14,000 g, 2 min twice. Catch tubes were replaced with new tubes. 50 µl of DNase/Pyrogen-free water were added to each tube and resuspended gently. Tubes were centrifuged at 14,000 g, 1 min. DNA in catch tubes were stored at -20℃. PCR was performed with Taq DNA polymerase and primers to the V1 - V3 region of the 16S rRNA gene. Then, the amplified DNA was used as template for 454 GS Junior (Roche Diagnosis, Indianapolis, IN) pyrosequencing. Filter-passed 3,000 reads were subjected to operational taxonomic unit (OTU) analysis with the cutoff similarity of 97% identity by using CLcommunity™ softwere (14).

Gut microbiota direct innate immune cell development via promoting hematopoiesis (15). In addition, Rag1^-/- mice that lack all mature lymphocytes have different compositions of gut bacterial community compared to wild type mice (1617). To check the number and proportion of the HSPCs in response to change of commensal microbial composition via a lymphopenia, we measured population of the bone marrow HSPCs of C57BL/6 female wild type and Rag1^-/- mice. The absolute number of total bone marrow cells was decreased in Rag1^-/- mice than wild type (Fig. 1A). As assessed by flow cytometry, the absolute number and proportions of LSK cells (Lineage^- Sca-1⁺ c-Kit⁺), LT-HSCs (Long-term HSCs; Lineage^- Sca-1⁺ c-Kit⁺ CD150⁺ CD48^-), and ST-HSCs (Short-term HSCs; Lineage^- Sca-1⁺ c-Kit⁺ CD150⁺ CD48⁺) were dramatically reduced (Fig. 1B), while MPs (myeloid progenitors; Lineage^- Sca-1^- c-Kit⁺) were increased in the bone marrow of Rag1^-/- mice (Fig. 1C and D). Similarly, percentage of ST-HSCs among LSK cells was reduced in the bone marrow of Rag1^-/- mice (Fig. 1E). On the contrary, percentage of MPPs (multipotent progenitors; Lineage^- Sca-1⁺ c-Kit⁺ CD150^-CD48⁺) was relatively up-regulated among LSK cells. Taken together, lymphopenic condition reduces the number of LT-HSCs and ST-HSCs and raises the number of MPs within the bone marrow.

Gut microbial diversity in co-housed animals is equilibrated by coprophagy (18). To examine homeostatic change of the bone marrow HSPCs according to gut microflora alteration, we used an additional bacterial administration to mimic recolonization (Fig. 2A). We extracted gut bacteria of feces from the wild type mice, and then orally injected the bacterial extraction of the feces to Rag1^-/- mice 4 times for 4 weeks. Another mice group was injected with the same volume of PBS as the bacterial injected Rag1^-/- mice. As expected, oral gavage treatment of feces increase the absolute number of total bone marrow cells of the Rag1^-/- mice (Fig. 2B). It also showed that the absolute number of LSK cells, LT-HSCs, ST-HSCs and MPPs were elevated as the level of wild types (Fig. 2C). Moreover, percentage of LSK cells, MPs, and MPPs among total bone marrow cells in gavaged Rag1^-/- mice were increased than the wild types (Fig. 2D). Especially, a proportion of ST-HSCs in feces-treated Rag1^-/- mice were significantly increased (Fig. 2D). Taken together, compositional change of gut microbiota in Rag1^-/- mice via feces transference affects homeostasis of HSPCs within the bone marrow.

To determine whether bacterial transfer induces compositional change of gut microbial community, we obtained bacterial DNA sample from feces of the wild type and Rag1^-/- mice. Bacterial compositions of the feces were analyzed by metagenomics analysis (Fig. 3A). It was shown that members of phylum Firmicutes were relatively less abundant and members of phylum Bacteriodetes had larger proportion in the feces-treated Rag-1-deficient mice than the PBS-treated Rag1^-/- mice (Fig. 3B). Family Lachnospiraceae in phylum Firmicutes were mostly reduced and family Helicobacteraceae in phylum Proteobacteria and family Prevotellaceae in phylum Bacteroidetes were relatively increased (Fig. 3B). Next, we tried to find a species that had the most significant changed of the three phyla. We focused on family Helicobacteraceae that had only identified bacterial species among the most abundant top 15 species in total microbiota of the feces-treated Rag-1-deficient mice (Fig. 3C). These data suggest that gut microbiota composition in Rag1^-/- mice could be altered by the injection of wild type mice gut bacteria.

In our metagenomics results, we found two species of bacteria, such as Helicobacter muridarum (H. muridarum) and Helicobacter ganmani (H. ganmani), which were abundant and identified. Since other bacterial species were unidentified or reduced in the feces-treated Rag1^-/- mice, we focused on two Helicobacter bacteria. To examine whether these species of bacteria affect homeostasis of the host HSPCs in the bone marrow, we treated a single-species bacterial injection experiment using male C57BL/6 Rag1^-/- mice. One group (5 mice) had been treated with H. muridarum via oral gavage 3 times for 4 weeks, while the other group (3 mice) had been treated PBS. Both groups were sacrificed at day 28 after first administration (Fig. 4A). Although absolute number of total bone marrow cells were significantly increased in bacteria-treated mice (Fig. 4B), the absolute number and proportions of HSPCs in the H. muridarum-treated mice were not up-regulated (Fig. 4C and D). The proportion of LT-HSCs, ST-HSCs, and MPPs among LSK cells had not shown significant change (Fig. 4E). These suggest that a single-species bacteria transfer of H. muridarum cannot induce the change in HSPCs homeostasis in the bone marrow in Rag-1-deficient mice.