Commensal Bacteria Are Dispensable in Neutropenia-Induced G-CSF Regulation

G-CSF is generally accepted to be the major cytokine regulating neutrophil production, but neutrophil steady-state homeostasis has still not been fully elucidated. We have previously shown in NODSCID^cγ−/− mice that feedback G-CSF expression upon antibody-based neutrophil depletion occurs independent from lymphocytes (Bugl et al., ASH 2010) and hypothesized that there may be a sensing mechanism of neutropenia relying on the presence of microbial components, similar to emergency granulopoiesis. Therefore, germ-free wildtype mice received anti-neutrophil antibodies and underwent further analysis. Moreover, passive regulation of circulating G-CSF levels by neutrophil cell mass was examined.

Anti-Ly6-G antibody (clone 1A8) was used to induce neutropenia in germ-free C57BL/6 mice. After one week of neutrophil depletion hematopoietic tissues and peripheral blood were harvested and analyzed on cellular, protein and RNA level. Moreover, neutrophil granulocytes (granulocyte-differentiation antigen-1⁺ Mac-1⁺) purified from peripheral blood of C57BL/6 mice were transfused into neutropenic wild type mice and plasma G-CSF was monitored.

Peripheral blood neutropenia could be effectively induced in all experimental mice by anti-1A8 antibody. Transfusion of 3.5 × 10⁶ neutrophils into neutropenic wild type mice did not significantly change plasma G-CSF levels. Filgrastim (rhG-CSF), however, caused significant downregulation of bone marrow G-CSF at the mRNA level. Germ-free C57BL/6 mice were analyzed after 1 week of antibody-induced neutropenia and showed a shift of myeloid progenitors towards granulocyte macrophage precursors (GMP) at the expense of megakaryocyte erythrocyte progenitors (MEP) as well as significantly increased numbers of hematopoietic stem cells. In addition, G-CSF, M-CSF, and CXCL12 behaved identically in both germ-free control and wild type mice under specific-pathogen free conditions.

Transfusion of G-CSF-receptor (CSF3R) positive neutrophils did not significantly influence G-CSF. Moreover, exogenous G-CSF downregulated marrow G-CSF on the transcriptional level. Germ-free mice are able to mount a feedback loop including G-CSF upregulation and marrow changes in progenitor cell distribution analogous to mice carrying physiological commensal bacteria. Our data are therefore consistent with G-CSF feedback regulation occurring independent from commensal germs. Moreover, our data indicate transcriptional rather than CSF3R⁺ cell mass-associated passive regulation of G-CSF levels. Taken together, we propose a model of myeloid bone marrow homeostasis, where feedback loops of neutrophil production upon antibody-dependent neutropenia occur through transcriptional upregulation of G-CSF. The underlying mechanisms occur independent of lymphocytes and presence of germs.

No relevant conflicts of interest to declare.