Mesenchymal stem cells (MSC) can be derived selectively in culture from multiple organs, an omnipresence we have recently suggested to be explained by the perivascular location of native MSC ancestors within intact tissues (Crisan et al. 2008, in press). We have now analyzed the ability of MSC extracted pro- or retrospectively from different human tissues to support hematopoiesis. MSC were either classically derived in primary cultures of umbilical cord blood (UCB) lineage-depleted mononuclear cells (n=3) or enzymatically dissociated adult adipose tissue (n=3), or grown as CD146+ NG2+ CD34-CD56- CD45- pericytes (n=2) purified by flow cytometry from fetal skeletal muscle and cultured over the long term. In both settings, identical MSC were obtained that maintained a stable CD146+ CD90+ CD73+ CD105+ CD34- CD45- surface phenotype and could differentiate into skeletal muscle, fat, bone and cartilage. CD34+ hematopoietic progenitors (n=3) immunoselected from term UCB were seeded (5×10e3cells/cm2 in triplicate) onto confluent irradiated layers of MSC derived from UCB, adipose tissue or fetal muscle pericytes (MSCu, MSCa and MSCmp, respectively) or, as a control, MS5 bone marrow stromal cells that allow the proliferation of very primitive human progenitor cells. All studies were approved by the relevant institutional regulatory board. The cells were cocultured for 5 weeks in a classical long-term culture-initiating cell assay in a complete medium (MyeloCult H5100, Stem Cell Technologies) containing hydrocortisone but no added cytokine. Wells were scored daily for the presence of cobblestone areas (CA) and half of the medium was replaced every week. Eventually, trypsinized cells from each well were characterized by flow cytometry for the expression of hematopoietic cell markers and assayed for CFC potential. After 14 days of incubation, colonies grown in semi-solid medium were scored as derived from colony forming units (CFU)-granulocyte, erythroid, macrophage, megakaryocyte (GEMM) and as high-proliferative-potential colony precursors (HPPC), the most primitive hematopoietic cell so far identified in a clonogenic assay in vitro. Within the CD45+ gate, all trypsinized cultures contained comparable percentages of CD34+lin- cells (MSCu: 51±9%; MSCa: 58±14%; MSCmp: 61±19%; MS5: 59±18%), the most immature hematopoietic cell compartment maintained during the long-term coculture. MSCu and MSCmp supported a similar cell proliferation during the whole culture while on MSCa, CA formed very rapidly and consistently but eventually decreased over the long-term culture. Interestingly, MSCu and MSCmp supported the development of the highest numbers of HPPC and of CFU giving rise to the largest GEMM colonies, as compared to MSCa that gave the same results as the control MS5 cell line. In summary, all MSCs tested were able to support hematopoiesis and CA formation, albeit with differences in growth kinetics and morphology of the colonies. Herein we show for the first time that purified human perivascular cells exhibit robust hematopoiesis support in vitro, in addition to multilineage mesodermal developmental potential. In conclusion, we demonstrate that MSC from novel sources distinct from the bone marrow are able to support hematopoiesis. These results further sustain the identity, beyond acronyms, between marrow stromal cells, long known for their support of hematopoiesis, and mesenchymal stem cells that gained more recent credit in the field of regenerative medicine because of their multilineage differentiation potential.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author