Development of three-dimensional models for the study of the hematopoietic stem cell niche

Most information about hematopoietic stem cell (HSC) self-renewal and multilineage commitment derives from standard bi-dimensional (2D) cell cultures or animal model systems. Through these approaches many important improvements in the understanding of the haematopoietic development have been achieved. However, none of the in vitro systems is capable of sustaining the haematopoiesis for a long time, whereas animal models may not adequately reproduce the features of human pathological conditions. Recently, several experiments using three-dimensional (3D) artificial systems have been carried out to expand haematopoietic progenitors and to achieve multilineage haematopoietic differentiation. The main aim of our study was to develop a new in vitro three-dimensional (3D) model of BM by employing natural components, to reproduce as close as possible BM microenvironment and the structure of HSC niches. To this aim, we employed human bone mineral matrix as scaffold and mesenchymal stem cells (MSCs) to reproduce the stromal microenvironment. Three different culture conditions were compared, aimed to achieve osteoblast, adipocyte and endothelial differentiation, respectively. After 2 weeks of CD34+ cell culture in the 3D system, immunohistochemistry as well as CFC and LTC-IC assays were carried out to investigate the system ability to maintain and differentiate HSCs. The exposure of MSCs to angiogenic stimuli in the presence of placental derived growth factor (PlGF) could reproduce, at least in part, the properties that make the microenvironment capable of supporting haematopoiesis. CD34+ cells grown in this 3D system differentiated into myeloid cells (CD45+/CD33+), but some of them remained in an undifferentiated status, as shown by the presence of CFC and LTC-IC after the culture. Immunohistochemical and gene expression analysis, carried out on the 3D system upon the angiogenic conditions before CD34+ cell culture, excluded the acquisition by MSCs of the endothelial-like phenotype, as well as the vascular smooth muscle cell differentiation. Instead, quite unexpectedly, the up-regulation of some of the early osteoblastic markers, together with angiopoietin 1 (ANGPT-1), suggested that the angiogenic conditions could have induced the differentiation of MSCs into HSC-supporting osteoblastic progenitors. In conclusion, our findings suggest that this new 3D model, based on the employment of MSCs and a natural scaffold of trabecular bone, is capable of inducing survival and differentiation of HSCs, probably through the creation of a osteoblastic-like HSC niche. Although further improvements, standardization processes, and mechanism evaluation are needed, this new tool may be useful to reproduce in vitro the conditions for HSC maintenance, expansion and differentiation, thus permitting to study the mechanisms of BM haematopoiesis in normal conditions and in haematological diseases.