Bone marrow is a complex organ system composed of two distinct lineages of cells: the hematopoietic cells and the supporting stromal cells, often referred as hematopoietic microenvironment (HME). Mesenchymal stem cells (MSCs) in bone marrow are shown to give rise to some of the components of HME, including osteoblasts, adipocytes and stromal fibroblasts in vitro, and to endothelial cells in vivo. It is a well accepted, but not definitely proven, concept that the HME provides structural niches, where dormant hematopoietic stem cells (HSCs) reside, and controls their renewal and differentiation. Although cotransplantation of human MSCs together with human HSCs resulted in increased chimerism of HSCs in animal models, existence of donor MSCs could only be detected using sensitive PCR-based analysis. Until this date, there is no physical evidence that transplanted MSCs have indeed engrafted in bone marrow and directly participated in that biological effect. In this study, we present the visual evidence for the sustained integration of human MSCs in murine bone marrow. Furthermore, we are able to delineate the physical interaction of injected human MSCs and cord blood derived CD34-positive HSCs (CBCD34).

In order to assess the spatial distribution, lineage commitment and interaction of MSCs and HSCs in situ, we transplanted green fluorescent protein (GFP)-transduced MSCs and yellow fluorescent protein (YFP)-transduced CBCD34 into tibia of NOD/SCID mice. Ten weeks after intramedullary injection, longitudinal sections of mouse tibiae were made and stained with various antibodies for multicolor immunofluorescent analysis using a confocal microscope. We detected not only the existence of GFP-expressing MSCs in bone marrow, but also differentiation into several cell lineages. GFP-expressing cells exhibited phenotype and morphplogy of N-cadherin-positive bone lining osteoblasts, osteocalcin-positive osteocytes in bone, cells lining abluminal surface of vasculature, and in rare occasion, CD34 and CD31-positive endothelial cells. We then quantitatively evaluated the proportion of GFP-MSCs interacted with primitive YFP-CD34 and lineage committed YFP-CD15 and -Glycophorin-expressing cells as well as the proportion of above mentioned hematopoietic cells interacted with GFP-MSC. Approximately 50% of MSCs associated with CD34-posititive stem cells compared to only 2% and 3% of those with CD15 and Glycophorin-positive cells, respectively. It was also evident that the frequency of CD34-positive cells interacted with MSCs was significantly higher than those with CD15 and Glycophorin-positive cells. The results were consistent with a long appreciated notion that more primitive cells closely interact with hematopoietic supporting stromal cells. Furthermore, we quantitatively proved that the majority of YFP-CD34-positive HSCs were found close proximity to the bone. By transplanting GFP-MSCs together with YFP-HSCs, this study provided direct visual evidence that transplanted human MSCs engrafted in murine bone marrow and integrated into HME, which physically interacted with human HSC.

Author notes

Corresponding author