The bone marrow microenvironment is a crucial component of hematopoietic stem and progenitor cell regulation, in both health and disease. Intercellular communication between hematopoietic and stromal cells is vital, for example, through the expression of cell surface or secreted factors, such as growth factors. Bone marrow stroma is composed of a multitude of heterogeneous cells, including those forming connective tissue and blood vessels, and those giving rise to a variety of supportive tissues including bone, fat, and cartilage. To date, it has been unclear whether the current cellular markers we use to identify stromal cells define truly distinct populations of cells. Definitive characterization of the cellular components of the hematopoietic niche is vital to understanding their interactions with hematopoietic cells in health and disease. Single-cell genomics is a powerful tool to characterize cellular architecture of cell populations such as the bone marrow niche. Dr. Ninib Baryawno and colleagues at Harvard University applied single-cell RNA sequencing to definitively annotate mouse bone marrow stromal cells based on their gene expression profiles. The authors define six overarching groups of stromal cells — mesenchymal stem cells (MSCs), osteolineage cells (OLCs), pericytes, chrondrocytes, bone marrow endothelial cells (BMECs), and fibroblasts. They further defined 17 distinct subpopulations, resolving many new subgroups and providing clarity on how certain populations are developmentally related.
First, the researchers were able to subclassify MSCs into four major subsets, distinguished by their expression of particular genes and elucidating likely pathways of differentiation between the subsets. The authors then enhanced the classification of what was previously defined as a single population of OLC into two subgroups with different origins and differing ability to regulate hematopoietic cells. With regard to cartilage formation, five different clusters expressed genes associated with the cartilage-forming lineage, and the authors used bioinformatic techniques to delineate potential differentiation pathways between the clusters of cells. To date, it has been difficult to distinguish mesenchymal stem cells from fibroblasts. In this study, however, the authors were able to recognize five different subsets of fibroblasts of varying likenesses to MSCs, from those expressing hematopoietic niche factors through to tendinous/ligamentous cells. Dr. Baryawno and colleagues also defined three distinct subsets of BMECs; all were related to each other along a continuum but express different hematopoietic ligands and secreted factors. They further described three pericyte subpopulations, also varying in their expression of key hematopoietic regulators such as CXCL12 and Kitl.
Following the mapping of bone marrow stromal cells in steady-state normal bone marrow, the authors went on to investigate the impact of acute myeloid leukemia (AML) on the stromal environment. They compared the effect of transplanting healthy mice with either normal bone marrow cells or MLL-AF9–driven leukemic bone marrow on the bone marrow stromal compartment. These studies demonstrated that AML distorted the stromal environment significantly, with significant changes in the proportions of subpopulations of stromal cells on exposure to leukemic bone marrow cells, including alterations in expression of hematopoietic regulators within defined stromal subpopulations. These stromal cell changes were consistent with a model in which leukemia creates an aberrant hematopoietic niche that simultaneously promotes aberrant leukemia cell differentiation and proliferation while suppressing normal hematopoiesis (e.g., via deregulation of vital hematopoietic stem cell niche growth factors, specifically Cxcl12 and Kitl).
In Brief
In summary, the authors provide novel insights into cellular heterogeneity of mouse bone marrow stromal cells, providing a comprehensive resource for researchers studying the bone marrow niche. They not only define novel subsets of stromal cells but also determine how these populations are related to one another along differentiation continuums and define expression patterns of key hematopoietic regulators. Importantly, this taxonomy was descriptive and almost entirely based on gene expression with manual annotation of putative functions. As the authors acknowledge, current methods do not allow in vivo functional assessment of each identified cell cluster. The presence of AML cells clearly disrupted the stromal environment, which is consistent with a model whereby leukemic cells influence the stromal cells, subverting their differentiation patterns and reducing the expression of regulatory signaling molecules known to be essential for normal hematopoietic function. The next challenge will be to map the spatial relationships between these identified stromal cell populations and hematopoietic or leukemic cell subpopulations using emerging imaging and spatial transcriptomic techniques.
Competing Interests
Dr. Chowdhury and Dr. Mead indicated no relevant conflicts of interest.
