Is normal hematopoiesis maintained solely by long-term multipotent stem cells?

Abstract

The understanding of the hierarchical organization of the human hematopoietic system is of major biologic and clinical significance. The validity of the conventional model in which hematopoiesis is solely maintained by a pool of multipotent long-term hematopoietic stem cells (LT-HSCs) has been recently challenged by several mouse studies. These new data point to the existence of a heterogeneous stem cell population that consists of distinct subsets of LT-HSCs, which include stem cells biased toward lineage-specific differentiation programs. This review attempts to discuss the balanced versus biased patterns of lineage output of human LT-HSCs gathered in 3 different gene therapy trials on the basis of vector integration site analysis by deep sequencing. The distribution of integration sites observed tends to support the validity of the revised model.

Figure 1

Homeostatic myeloid-biased cell expansion. In the human β-thalassemia trial, most of therapeutic effect results from a dominant myeloid-biased cell clone. In this clone, the HMGA2 integration site (IS) is present in similar proportions among erythroblasts, granulocyte, monocyte, and LTC-IC cells, but not in lymphocytes (LT); whereas its expression is strictly erythroid specific. Therefore, the HMGA2 IS-initiating cell is probably a myeloid-biased LT-HSC19/a cell with increased downstream cell production (1, in red) or a common myeloid progenitor with acquired self-renewal capability (2, in red). From Cavazzana-Calvo et al.