In Vivo Depletion of Osteoblasts Reveals Roles in HSC Homing and B-Lymphopoiesis.

The identification of bone marrow (BM) niches controlling many aspects of hematopoiesis is of great interest. It is believed that these niches provide the necessary signals to maintain primitive hematopoietic cells including hematopoietic stem cells (HSC). Downstream of HSC are programs for both myelopoiesis and lymphopoiesis, although little is known about the niches that support these processes. Accumulating data implicating osteoblasts (OB) as a component of the HSC niche suggest that these cells are able to regulate HSC properties, and that the HSC are located in direct contact with OB at the bone endosteum. However, the interpretation of these data is controversial, and others have suggested that endothelial cells dictate the structure and function of the HSC niche. In order to determine whether OB are a critical component of hematopoietic niches, we employed Col2.3ΔTK mice, a transgenic mouse strain bearing a fusion gene composed of the rat type I collagen promoter and HSV-thymidine kinase. This construct allows for selective depletion of OB following in vivo gancyclovir (GCV) treatment. Mice received 100 mg/kg GCV for a period of 8, 21, or 28 days. Following 8 days of GCV, OB were depleted from the BM as determined by a lack of staining for osteocalcin. Concommitant with the loss of OB, there was a decrease in total BM cellularity of approximately 50% that included reductions in both myeloid and B lymphoid cells. Among the B cell populations that were measured, pre-pro-B and pro-B cells were reduced to the greatest extent, 64% and 81%, respectively. Despite this marked reduction in B lymphopoiesis, number of Lin− Sca-1+ cKit+ cells (LSK), a population enriched for HSCs, was unchanged immediately following OB depletion. This loss of B cell precursors, occurring before any observable effects on HSC numbers, suggests that OB may directly support the proliferation and maturation of early B cell precursors. Although minimal changes in BM LSK were observed immediately after GCV, BM LSK were reduced after 21 and 28 days of GCV treatment, when their numbers decreased 40% and 88%, respectively. During this same period, however, spleen LSK increased by 165% and 300%, respectively. Assuming that femurs and tibiae constitute 25% of total murine BM, the total number of LSK lost from the BM after 21 and 28 days of GCV is approximately 1.42 and 3.12×10^5 cells, respectively. These values are remarkably consistent with the increase in splenic LSK at the same time points, 1.2 and 3.7×10^5 cells, respectively. These data indicate that loss of OB plays a pivotal role in the partitioning of LSK into and out of the BM. Given the finding that the shift in LSK is most pronounced 2–3 weeks after the loss of OB, these data raise the intriguing possibility that the partitioning of LSK from the BM to the spleen following OB deletion is not the result of increased emigration from the BM but, rather, due to a decreased ability of HSC to home back to the BM following daily limited physiological mobilization. For example, it is plausible that OB supply chemotactic signals that control the ability of HSC to home to the BM. In conclusion, in vivo depletion of OBs causes an immediate reduction in the number of B cell precursors with a delayed shift in LSK from the BM to the periphery, thus demonstrating that OB play a major role in BM HSC trafficking and B lymphopoiesis.