The Bacterial Microbiota Promotes Normal Hematopoiesis Via Interferon Alpha and NOD1 Signaling Pathways

Long-term antibiotic therapy is associated with hematological side effects such as neutropenia and anemia. Our lab and others have shown that long-term antibiotic treatment in mice leads to bone marrow suppression and agranulocytosis in mice through depletion of the commensal bacteria. Our work further showed that Stat1-deficient mice phenocopy the bone marrow suppression phenotype of antibiotic-treated mice, suggesting that commensal microbiota mediate hematopoiesis via Stat1 signaling. However, the upstream actors of this pathway and the bacterial mediators required for commensal microbiota regulation of normal hematopoiesis still remain poorly understood. Such knowledge will be essential for understanding how to treat antibiotic-associated cytopenias. We hypothesize that microbial products detected by host cells trigger STAT1 signaling to potentiate normal hematopoiesis.

To identify the host cells that require Stat1 for microbiota-promoted hematopoiesis, we treated conditional Stat1 knock-out mice with two weeks of antibiotic therapy. Of the four conditional knock-out mice we evaluated (LepR-Cre, Villin-Cre, Vav-iCre, LysM-Cre), only the mice deficient in STAT1 in hematopoietic cells (Vav-iCre Stat1 ^fl/fl) phenocopied the bone marrow suppression of antibiotic-treated mice. Our data suggest that STAT1 signaling is necessary in non-myeloid hematopoietic cells, but not intestinal epithelial cells or mesenchymal stromal cells for microbiota-promoted hematopoiesis. Non-competitive transplantation of Stat1 ^-/- bone marrow into wild type mice validated these findings; mice lacking STAT1 only in hematopoietic tissues phenocopied the Vav-iCre mice, consistent with a specific role for microbiota-mediated STAT1 signaling in the hematopoietic compartment.

To assess the upstream mediator of STAT1 signaling in this biological context, we treated interferon (IFN) receptor knock out mice with two weeks of antibiotics. Of the three types of IFN receptor knock-outs evaluated, only mice deficient in type I IFN signaling phenocopied the bone marrow suppression of antibiotic-treated mice. These findings suggest that type I IFN signaling, and not type II or III, was required for microbiota-dependent hematopoiesis. These results were validated by showing that the administration of pegylated-IFNα was sufficient to rescue the depletion of type I IFN-STAT1 signaling in antibiotic-treated mice.

To determine the microbial signals that may potentiate hematopoiesis, we evaluated two microbial products that were previously shown to alter hematopoiesis and to activate type I IFN signaling (Iwamura et al. Blood 2017 & Steed et al. Science 2017). We discovered that oral administration of these commensal-derived products, the metabolite desaminotyrosine (DAT) or NOD1 ligand (NOD1L), a motif of peptidoglycan, were each sufficient to rescue the hematopoietic defects induced by antibiotics in mice. To test whether these products rescue hematopoiesis by activating STAT1 signaling, we attempted to rescue the hematopoietic defects in Stat1 ^-/- mice. These studies showed that NOD1L rescues granulocyte but not progenitor counts in Stat1 ^-/- mice, suggesting that NOD1 and type I IFN signaling work together at the progenitor level, but independently at the downstream myeloid progenitor level to promote granulopoiesis.

Overall, our studies expand our understanding of the signaling pathways by which the microbiota promotes normal hematopoiesis and identify novel therapeutic agents that can be used to ameliorate antibiotic-induced BM suppression.