Hematopoietic stem cells (HSC) have been successfully used to treat leukemias, inherited immune deficiencies and other life-threatening blood diseases 
. However, only a fraction of patients benefit from this therapy due to the lack of HLA-matched bone marrow donors, and low number of HSC in cord blood 
. Therefore, a long-standing goal has been to establish culture protocols to facilitate HSC expansion. However, there has been little success in expanding human HSC for clinical purposes due to limited understanding of the complex mechanisms governing HSC properties, and how these programs become compromised in culture. Furthermore, most HSC regulators have been identified using gene-targeted mouse models 
, whereas mechanistic understanding of human hematopoiesis is lagging behind due to lack of suitable in vitro
and in vivo
model systems for manipulating human HSC or their niche. A major challenge in culturing HSC is the difficulty to recreate the specialized microenvironment that regulates self-renewal of HSC within hematopoietic tissues; as a result, cultured HSC are subjected to rapid differentiation or death 
The bone marrow HSC niche consists of multiple cell types, including mesenchymal stem cells (MSC), osteoblasts, adipocytes, endothelial cells and macrophages 
. The microenvironment directs HSC fate decisions by mediating cell-cell interactions and secreting soluble growth factors 
. Although several HSC supportive cytokines (e.g. SCF, IL-11, IL-3, FLT-3, TPO, angiopoietin-like proteins, and the Notch1 ligand Dl1) 
, cell-intrinsic stimulators of HSC expansion (e.g. HOXB4) 
and inhibitors of negative HSC regulators (e.g. AhR signaling 
) have been identified, these have not yet led to the establishment of routine clinical protocols for HSC expansion. Several studies have assessed the suitability of various stromal cell lines from fetal and adult hematopoietic tissues to support murine and human hematopoiesis 
; nevertheless, there has been little progress in expanding functional human HSC on these stroma lines. It is unclear to what extent the different HSC properties can be maintained in culture, and what molecular defects prevent robust expansion of transplantable HSC. Understanding how the ex vivo
culture per se
affects HSC function and molecular properties will be a critical step toward improving culture conditions for the expansion of HSC for clinical purposes, and also for the long-term goal to generate transplantable HSC in culture from human pluripotent stem cells.
To understand the behavior of human hematopoietic stem/progenitor cells (HSPC) in culture we established an MSC stroma based co-culture system for modeling human hematopoietic hierarchy, and defined the extent to which surface markers, functional properties and transcriptome characteristic for the primitive HSPC fraction can be preserved during culture. We show that OP9M2, a subclone of OP9 stroma cells, protects human fetal liver and cord blood HSPC from differentiation and apoptosis, facilitating a dramatic ex vivo expansion of multipotent hematopoietic cells that preserve the CD34+CD38−CD90+ surface immunophenotype that is characteristic for human HSC. This system also maintains the initial number of transplantable human fetal liver HSC (defined based on myelo-lymphoid reconstitution in NSG mice) for at least 2 weeks in culture, but does not support their significant expansion. Genome-wide gene expression analysis of the expanded fetal liver CD34+CD38−CD90+ cells showed a remarkably stable transcription factor network associated with HSC entity, but revealed dynamic changes in distinct molecular programs that are sufficient to compromise HSC function. Thus, this co-culture offers a robust ex vivo system for studying the regulation of human multilineage hematopoiesis. Furthermore, the temporal gene expression data from in vivo derived and ex vivo expanded human CD34+CD38−CD90+ will serve as a resource to identify key regulatory mechanisms that control HSC identity vs. function, and to develop clinically applicable protocols for HSC expansion and de novo generation from pluripotent stem cells.