Abstract 721

Hematopoietic stem/progenitor cells (HSPCs) are released from the bone marrow (BM) to the circulation by granulocyte-colony stimulating factor (G-CSF) via sympathetic nervous system (SNS)-mediated osteoblast suppression (Katayama et al. Cell 2006). We further elucidated that vitamin D receptor is essential for this neuronal control of endosteal niche (Kawamori et al. Blood 2010). Osteoblasts are known to adopt three fates: die by apoptosis, become bone-lining cells, or become embedded in osteoid and then in mineralized bone matrix to terminally differentiate into osteocytes, which constitute more than 95% of bone cells. Osteocytes have been shown to control the functional balance between osteoblast and osteoclast via mechanotransduction. In order to address the role of bone-embedded osteocytes in HSPCs niche function, we first quantified mRNA expression of bone-related genes in the femur of wild-type (WT) mice to examine if osteocytic function changes during G-CSF treatment (125μg/kg/dose, 8 divided doses, every 12 hours). Whereas markers relating to osteoblast function, osteocalcin and osteopontin, started to decrease late at 6 doses of G-CSF administration when mild mobilization of HSPCs had occurred, osteocyte-specific genes, including neuropeptide y, SOST, MEPE, E11/gp38 and Phex, were rapidly suppressed at 1 dose when no mobilization was observed. These data suggest that osteocytes respond to G-CSF with altered gene expression much earlier than osteoblasts. Further, the number and thickness of osteocyte projections extending toward the endosteal surface were markedly reduced, as assessed by fluorescently labeled phalloidin, at 8 doses of G-CSF treatment when full mobilization was achieved; these morphological changes were observed specifically in newly-embedded osteoid osteocytes, but not in mature osteocytes embedded deep inside mineralized bone. These findings suggest that osteoid osteocytes may sense the signal triggered by G-CSF. We confirmed the presence of β2-adrenergic receptor in osteoid osteocytes and tyrosine hydroxylase-positive nerve fibers in the vicinity by immunofluorecence staining, suggesting that osteoid osteocytes may be regulated by SNS. To directly address osteocyte involvement in G-CSF-induced mobilization, we utilized a transgenic (TG) mice in which inducible and specific ablation of osteocytes is achieved through targeted expression of diphtheria toxin (DT) receptor under DMP-1 promoter. A single injection of DT in TG mice generates “osteocyte-less (OL)” mice. We found that mobilization by G-CSF was drastically impaired in OL mice for progenitors (CFU-Cs, mean±SEM, WT vs Tg: 1673±271 vs 242±94/ml blood, n=6-13, p<0.01; lineage-Sca-1+c-kit+ (LSK) cells, WT vs Tg: 6878±1209/ml vs 1763±502/ml, n=6-13, p<0.01) and stem cells (repopulating units at 4 months, WT vs Tg: 2.5±0.7 vs 0.5±0.2, n=6-7, p<0.05), while the OL BM showed normal HSPC number. The levels of CXCL12 mRNA and protein in BM and bone were markedly decreased during G-CSF treatment even in OL mice despite the mobilization defect, and a CXCR4 antagonist AMD3100 induced mobilization normally in the absence of osteocytes. Thus, osteocytes embedded within the bone are indispensable for G-CSF-induced mobilization through a CXCL12-independent mechanism. Although most of bone-related genes exhibited drastic decreases following G-CSF treatment, we found that fibroblast growth factor 23 (fgf23) mRNA displayed a 4-fold increase at 6 doses of G-CSF. FGF23 is mainly produced by osteocytes and Klotho is an obligate coreceptor for FGF23 to bind and activate FGF receptors. Since we confirmed that klotho hypomorphic (kl/kl) mice showed remarkably disrupted osteocyte network, we injected G-CSF into these mice. As we expected, G-CSF induced virtually no mobilization in kl/kl mice while the number of HSPCs in the BM remained comparable to control mice. Collectively, our results demonstrate a novel function of bone-embedded osteocytes as a critical regulator of HSPC trafficking perhaps by controlling the endosteal niche and establish the important physiologic function of skeletal tissue for hematopoietic microenvironment.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution