Single-Cell RNA Sequencing Identifies Phenotypically, Functionally, and Anatomically Distinct Stromal Niche Populations in Human Bone Marrow

Hematopoiesis is regulated by the bone marrow (BM) stroma. However, cellular identities and functions of the different BM stromal elements in humans remain poorly defined. Based on single-cell RNA sequencing (scRNAseq) technology we herein investigated the human non-hematopoietic BM cell compartment (CD45^low/-CD235a^-) enriched with highly-enriched stromal cells (CD45^low/-CD235a^-CD271⁺) aiming to resolve the composition of the human HME at the highest possible resolution and to identify potential novel marrow stromal subsets and cellular hierarchies.

Our dataset comprised of 25067 cells which formed 42 clusters corresponding to distinct cell types and differentiation stages. Twelve non-hematopoietic clusters including stromal cells could be clearly identified. Based on their transcriptomic signatures, stromal clusters were annotated into six different stromal cell types: multipotent stromal stem cells (MSSCs), adipo-primed progenitors, balanced progenitors, pre-osteoblasts, osteochondrogenic progenitors (OCs), and pre-fibroblasts.

RNA velocity analysis allowed us to recapitulate stromal cell differentiation pathways and established that the MSSC cluster was at the apex of a differentiation hierarchy with downstream differentiation paths into adipo-primed progenitors, balanced progenitors, OCs and pre-osteoblasts. RNA velocity analysis also allowed us to detect potential key regulatory factors, including putative driver genes that govern cell fate commitment (i.e. CEBPD for adipogenesis and VCAN for osteogenesis).

Based on the differential expression of NCAM1, CD52 and CD81, we then designed a fluorescent-activated cell sorting (FACS) strategy to isolate the different stromal cell types. Functional analysis of sorted stromal cells showed that both MSSCs and adipo-primed progenitors demonstrated considerably higher in vitro colony-forming capacities compared to other stromal cell types. MSSCs and adipo-primed progenitors exhibited full multi-differentiation potentials, whereas pre-fibroblasts had only limited differentiation potentials. OCs showed enhanced osteoblastic and chondrogenic differentiation capacities but compromised adipogenic potential, which was consistent with expression profiles.

In-situ staining using human BM biopsies demonstrated that CD271/CD81 double-positive cells (i.e. MSSCs and adipo-primed progenitors) were localized in BM stromal regions while CD271/NCAM1 double-positive cells (i.e. OCs) were exclusively localized endosteally. These results thus demonstrated distinct anatomical localizations of MSSCs and OCs, which confirms and extends our previous findings on differently localized hematopoietic niche cells.

Finally, cell-cell communication analysis showed that all stromal clusters demonstrated potential to interact with a wide range of hematopoietic cells including hematopoietic stem and progenitor cells (HSPCs). According to the prediction, MSSCs communicate with hematopoietic cells mainly through CXCL12-, VCAM1-, FN- and MDK-involving pathways while OCs interact with hematopoietic cells mainly via SPP1-mediated pathways. On the other side, stromal cells were regulated by hematopoietic cells through PDGF-, OSM-, and TGFB1-mediated signaling. Additionally, multiple intra- and inter-stromal pathways were identified, thus predicting a complex network of interactions in the human BM microenvironment.

In summary, we identified distinct, functionally different and hierarchically organized stromal stem and progenitor cell populations. Furthermore, our findings conceptually support a model where hematopoietic cells are differentially regulated by distinct niche milieus which are composed of specific cytokine-producing stromal cells, thus refining our current view on hematopoietic niche organization.

No relevant conflicts of interest to declare.