Bone forming osteoblasts an are essential component of bone marrow (BM) niches necessary for the maintenance of hematopoietic stem cells (HSC) and B-cell progenitors. Previous studies from our and other groups have shown that the mobilizing agent granulocyte colony-stimulating factor (G-CSF) reduces the number and function of osteoblasts lining the bone at the endosteum. By histomorphometry, we now demonstrate that like G-CSF, the chemotherapy agent cyclophosphamide (CY) also induces a rapid and dramatic reduction in osteoblast number and osteoid surface at days 6–8 when hematopoietic stem and progenitor cells (HSPC) are mobilized into the blood. Both G-CSF- and CY-induced HSPC mobilizations coincide with a 90% reduction in bone mineralization and formation as measured by incorporation of calcein, a Ca2+-chelating fluorophore, in the mineralizing bone. As osteoblasts form specific niches for HSC and primitive B lymphoid progenitors, we measured expression of hematopoietic and lymphopoietic cytokines at the endosteum to gain insight on how these endosteal niches are affected during HSPC mobilization. As expected, osteoblast and osteoprogenitor specific mRNA were markedly reduced during mobilization induced by G-CSF or CY with significant reductions in Runx2 (5- fold), osterix (12–32-fold), PTHR1 (20-fold) and osteocalcin (50–100-fold) mRNA. To determine whether this reduction in osteoblast number/function affected the expression of cytokines essential for the maintenance of HSC and B progenitor at the endosteum, we measured mRNA encoding CXCL12, IL-7, BAFF, KIT ligand and angiopoietin-1 (Ang-1) in total endosteal cells as well as in endothelial cells (Lin−CD45−CD31+), mesenchymal osteoprogenitors (Lin−CD45−CD31−Sca-1+CD51+) and osteoblasts (Lin−CD45− CD31−Sca- 1−CD51+) sorted from the endosteum. In saline-treated mice, IL-7 and CXCL12 were mostly expressed by osteoblasts, BAFF mostly by osteoprogenitors, Ang-1 by both osteoblasts and osteoprogenitors, whereas KIT ligand was expressed by both osteoblasts and endothelial cells. G-CSF caused a 3-fold reduction in the expression of IL-7, BAFF, and KIT ligand, a 4-fold reduction in Ang-1, and a 5–10-fold reduction in CXCL12 mRNA at the endosteum, demonstrating an impairment of endosteal niches during G-CSF administration. Despite maintenance of lineage-negative (Lin−) Sca-1+ KIT+ cell number in the BM, Lin− Sca-1+ KIT+ CD41− CD48− CD150+ phenotypic HSC were decreased 3.4- fold in mobilized BM. By measuring in vivo Hoechst33342 incorporation following a 10 minute perfusion, we found that these phenotypic HSC relocated from niches distal from BM blood sinuses (low Ho33342 incorporation in vivo) to more perivascular locations (high Ho33342 incorporation). Finally, the numbers of B progenitors were dramatically reduced in mobilized BM with enhanced B cell apoptosis. Mobilization of B progenitors into blood and spleen did not compensate for their reduction in the BM, resulting in a net loss in all B cell progenitors in whole mice. From our findings, we conclude that reduced osteoblast number/function during G-CSF-induced HSC mobilization not only leads to the impairment of HSC endosteal niches but also an arrest in medullar B lymphopoiesis.
Disclosures: No relevant conflicts of interest to declare.