Maintenance of neutrophil homeostasis in the blood is vital to the proper functioning of the innate immune system. Neutrophil release from the bone marrow is a major regulated determinant of neutrophil homeostasis in the blood, yet the molecular signals that regulate this process are largely unknown. Accumulating evidence suggests that stromal derived factor-1 (SDF-1), through interaction with its major receptor CXCR4, provides a key retention signal for neutrophils in the bone marrow. Definitive proof of this hypothesis has been hampered by the embryonic lethality of CXCR4−/ − mice and the severe engraftment defect observed in recipients of CXCR4−/ − fetal liver hematopoietic stem cells.

To circumvent this problem, in the present study we generated gene-targeted mice in which CXCR4 was selectively inactivated only in myeloid cells. The mice (CXCR4f/−LysM+/Cre) expressed Cre-recombinase under the control of the lysozyme-M promoter and contained one CXCR4-null allele while the other was flanked by loxP sites. Myeloid-specific loss of cell-surface CXCR4 expression was documented by flow cytometry. CXCR4 expression in other lineages was documented at levels comparable to controls. Peripheral blood counts in CXCR4f/−LysM+/Cre mice were normal except for marked neutrophilia. The absolute neutrophil count was 1.23 ± 0.76 and 8.05 ± 3.00 in wild type (wt) and CXCR4f/−LysM+/Cre mice, respectively (p <0.001). Of note, no increase in circulating hematopoietic progenitors or immature myeloid cells was observed in the peripheral blood of CXCR4f/−LysM+/Cre mice. In addition, no perturbation in B or T lymphocytes was detected. In the bone marrow, the number of Gr-1+ myeloid cells was reduced to 69.2 ± 24.1% of control mice. No accumulation of granulocytic precursors was observed, suggesting that granulocytic differentiation in CXCR4f/−LysM+/Cre mice was normal. As a metric for quantifying neutrophil distribution, we calculated the percentage of total body neutrophils in the blood (neutrophil distribution index or NDI) as described previously (Immunity, Vol. 17, 413–423, 2002). Consistent with previous reports, 1.0 ± 0.5% of neutrophils were in the blood of wt mice compared with 9.5 ± 4.2% in CXCR4f/−LysM+/Cre mice (p<0.001). Together, these data provide strong genetic evidence supporting the hypothesis that CXCR4 signals are key regulators of neutrophil release from the bone marrow.

A broad range of chemokines and cytokines are known to induce neutrophil release from the bone marrow. A recent report (Blood, Vol. 104, 565571, 2004) showed that treatment of neutrophils with KC resulted in heterologous desensitization of CXCR4. Another report (Blood, Vol. 108, 812820, 2006) demonstrated that G-CSF downregulates CXCR4. These observations raise the possibility that disruption of CXCR4 signaling may be a common mechanism by which all mobilizing agents induce neutrophil release. To test this hypothesis, we measured neutrophil mobilization one hour after a single 125 ug/kg injection of G-CSF, a time before any detectable change in SDF-1 expression levels in the bone marrow. Neutrophil numbers increased 1.92 ± 0.16 fold over baseline in wt mice (p<0.05). In contrast, no change was detected in CXCR4f/−LysM+/Cre mice (1.05 ± 0.39 fold over baseline, p=ns). These data show that neutrophil mobilization by G-CSF is dependent upon CXCR4. Studies are underway to characterize neutrophil mobilization by other mobilizing agents in CXCR4f/−LysM+/Cre mice.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author