Expression of the G-CSF receptor in monocytic cells is sufficient to mediate hematopoietic progenitor mobilization by G-CSF in mice

Abstract

Granulocyte colony-stimulating factor (G-CSF), the prototypical mobilizing cytokine, induces hematopoietic stem and progenitor cell (HSPC) mobilization from the bone marrow in a cell-nonautonomous fashion. This process is mediated, in part, through suppression of osteoblasts and disruption of CXCR4/CXCL12 signaling. The cellular targets of G-CSF that initiate the mobilization cascade have not been identified. We use mixed G-CSF receptor (G-CSFR)-deficient bone marrow chimeras to show that G-CSF-induced mobilization of HSPCs correlates poorly with the number of wild-type neutrophils. We generated transgenic mice in which expression of the G-CSFR is restricted to cells of the monocytic lineage. G-CSF-induced HSPC mobilization, osteoblast suppression, and inhibition of CXCL12 expression in the bone marrow of these transgenic mice are intact, demonstrating that G-CSFR signals in monocytic cells are sufficient to induce HSPC mobilization. Moreover, G-CSF treatment of wild-type mice is associated with marked loss of monocytic cells in the bone marrow. Finally, we show that bone marrow macrophages produce factors that support the growth and/or survival of osteoblasts in vitro. Together, these data suggest a model in which G-CSFR signals in bone marrow monocytic cells inhibit the production of trophic factors required for osteoblast lineage cell maintenance, ultimately leading to HSPC mobilization.

Figure 1.

Neutrophil number in the bone marrow correlates poorly with G-CSF–induced HSPC mobilization. Wild-type and Csf3r^−/− bone marrow cells were mixed at the indicated ratios and transplanted into irradiated recipients. 6 wk later, the mixed bone marrow chimeras were treated with G-CSF (250 µg/kg/d G-CSF for 7 d), and the number of CFU-C in the peripheral blood (A), spleen (B), or bone marrow (C) was measured. As a control, a cohort of mixed chimeras was analyzed without G-CSF treatment (UnRx). Because similar results were obtained with each type of chimera, the untreated data were pooled. The percentage of neutrophils in the bone marrow derived from wild-type cells (D) and the absolute number of wild-type neutrophils per femur (E) were determined after 7 d of G-CSF administration. (F) The number of CFU-C in the blood versus the number of wild-type neutrophils in bone marrow. The Pearson r² value is shown. Data represent the mean ± SEM of six to nine mice and is pooled from two independent transplantation experiments. *, P < 0.05 compared with untreated chimeras; ⁺, P < 0.05 compared with 3:1 chimeras.

Figure 2.

G-CSF–induced HSPC mobilization is normal in Rag1^−/− and NOD/scid/IL-2γ^null mice. Wild-type, Rag1^−/− (A), or NOD/scid/IL-2γ^null mice (B) were treated with G-CSF (250 µg/kg/d for 5 d) or left untreated. The number of CFU-C in the blood, bone marrow, and spleen was measured after 5 d of G-CSF. Data represent the mean ± SEM of four to five mice per genotype per treatment group and is pooled from four independent experiments. *, P < 0.01; **, P < 0.001.

Figure 3.

G-CSFR expression is mainly restricted to monocytic cells in CD68:G-CSFR mice. (A) Schematic of the CD68:G-CSFR transgene. IRES, internal ribosomal entry sequence. (B) Representative histograms showing biotinylated G-CSF binding in the absence (solid line) or presence (dashed line) of a 100-fold molar excess of unlabeled G-CSF in the indicated blood leukocyte population. G-CSFR surface expression is proportional to the difference in median fluorescence intensity (ΔMFI) between the two curves. (C) The mean ΔMFI in the indicated blood leukocyte population is shown. Data represent the mean ± SEM of four to five mice per genotype and is pooled from four independent experiments.

Figure 4.

G-CSF–induced HSPC mobilization is normal in CD68:G-CSFR mice. Mice of the indicated genotype were treated with G-CSF for 7 d or left untreated. The number of KLS (A) or CFU-C (B) in the blood, bone marrow, and spleen on day 7 is shown. The mRNA expression in the bone marrow on day 7 of osteocalcin (C) or CXCL12 (D) relative to β-actin is shown. Data represent the mean ± SEM of three to five mice per genotype per treatment group and is pooled from three independent experiments. *, P < 0.01 compared with untreated mice of the same genotype; **, P < 0.05 compared with untreated mice of the same genotype; ^†, P < 0.05; ^‡, P = 0.07 compared with G-CSF–treated Csf3r^+/− mice.

Figure 5.

G-CSF treatment leads to a loss of monocytic cells from the bone marrow. CX3CR1^GFP/+ mice were treated with PBS or G-CSF for 5–7 d and the flush and bone fractions harvested as described in Materials and methods. (A) Representative dot plot showing the gating strategy used to identify inflammatory (Gr-1^high GFP⁺) and resident monocytes/macrophages (Gr-1^low GFP⁺). (B) The absolute number of inflammatory and resident monocytes/macrophages in the flushed and bone fractions are shown. Data represent the mean ± SEM of four mice per cohort pooled from two independent experiments. *, P < 0.05; **, P < 0.005; ***, P < 0.001. (C) The percentage of inflammatory and resident monocytes in the flush fraction of bone marrow was determined at the indicated time points. In these experiments, wild-type mice rather than CX3CR1^GFP/+ mice were used, and CD115 was used to identify monocytic cells. Data represent mean ± SEM of two to six mice per cohort pooled from four independent experiments. **, P < 0.005 compared with untreated mice as determined by one-way ANOVA with Dunnett’s post-test to correct for multiple comparisons.

Figure 6.

Macrophages support the growth of mature osteoblasts in vitro. Unfractionated bone marrow cells from wild-type mice were cultured for 10 d and then sorted by flow cytometry into stromal (CD45⁻ Ter119⁻) and macrophage (CD45⁺ CD115⁺) cell populations. Stromal cells were cultured for an additional 14 d in the absence (No Mø) or presence (Mø) of an equal number of macrophages. Ascorbic acid and calcium were included in all cultures to stimulate mature osteoblast development. (A) Representative photomicrographs of cultures on day 14 that were stained for alkaline phosphatase. (B) Osteocalcin protein concentration in conditioned media was measured by ELISA at the indicated time points. (C and D) CXCL12 protein (C) and Osteocalcin (D) concentrations in conditioned media were measured by ELISA on day 14 of culture. Where indicated, the macrophages were separated from the stromal cells by a semipermeable membrane (transwell). Data represent the mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.005; ***, P < 0.001.