Combined Single-Cell and Spatial Transcriptomics to Deconvolute the Hematopoietic Stem Cell Niche

Coordinated interaction of many cell types is required to facilitate hematopoietic and mesenchymal stem cell maintenance and differentiation in the bone marrow. However, the molecular factors and cell types involved in this complex interplay remain poorly understood. Here we developed a combined single cell and spatial transcriptomics approach to address this problem.

Large-scale single-cell transcriptional profiling in conjunction with a multi-layered sorting approach allowed us to generate a complete and evenly sampled transcriptional map of all major bone and bone marrow populations. Our dataset covers all cell types or differentiation trajectories involved in mesenchymal and hematopoietic stem cell differentiation, osteogenesis, adipogenesis, myelopoiesis, erythropoiesis, lymphopoiesis, memory T cell formation as well as bone marrow neural innervation and vascularization at the single cell level. Using this data, we derive fundamental properties of the described cell types, clarify the cellular source of signals affecting stem cell differentiation processes and provide a systems view on putative intercellular interactions. Systematic spatial transcriptomics, using laser-capture microdissection of selected bone marrow niches followed by transcriptional profiling and bioinformatic cellular deconvolution, allowed us to confirm predicted interactions and map the cellular composition of distinct bone marrow niches.

Our analyses highlight the importance of pre-adipogenic CXCL12 abundant reticular cells as key niche cells for stem cell maintenance, provides a holistic systems view of the hematopoietic stem cell niche and offers a novel approach to systematically deconvolute the molecular, cellular and spatial composition of complex tissues.