34 + 43 = early human blood lineage

Comment on Vodyanik et al, page 2095

In this issue of Blood, Vodyanik and colleagues show that the CD43 antigen can be used to separate hematopoietic progenitors cells derived from human embryonic stem cells from other CD34+ expressing cells.

The isolation of human embryonic stem cells (hESCs) has created huge opportunities for the field of regenerative medicine, because hESCs can be directed to differentiate into many cell types. Some of the most clinically important cells that can be produced from hESCs are hematopoietic cells. Production of hematopoietic stem cells from hESCs, particularly from patient-specific ESCs, would provide an immensely valuable new source of cells for transplantation, because curative stem cell transplants are too often precluded by the lack of histocompatible cells. However, while production of hematopoietic cells from both mouse and human ESCs can easily be demonstrated using in vitro assays, repopulation of lethally irradiated mice with ESC-derived hematopoietic cells has remained difficult, and only modest engraftment can be achieved.^1,2  Better engraftment can be obtained by genetically modifying the ESCs,³  but genetically modified cells might not be suitable for clinical use. The difficulties in differentiating ESCs into transplantable hematopoietic cells could be due either to intrinsic defects in cells produced in culture or, more plausibly, to the fact that the cells produced from ESCs are at a very early developmental stage and are not functionally equivalent to adult hematopoietic stem cells.

Another promising avenue of research is the large-scale production of red blood cells (RBCs) for transfusion by differentiation of hESCs. While the current collection-based RBC supply is very safe, it remains vulnerable to emerging pathogens. Industrial production of RBCs from hESCs would eliminate the problem and yield cells of greater homogeneity. Importantly, since RBCs do not express human leukocyte antigens (HLAs) and are enucleated, two of the biggest issues in the hESC field, immune rejection and the possibility of oncogenic transformation, are greatly diminished. Recent work from the Douay lab in Paris has demonstrated that large numbers of apparently perfectly functional enucleated RBCs can be produced in culture by in vitro differentiation of human cord or peripheral hematopoietic stem cells,⁴  but RBCs produced from hESCs are probably not suitable for transfusion because they are similar to RBCs produced in the yolk sac and fetal liver early during human embryogenesis and express mostly embryonic and fetal globins.⁵ 

To solve these problems and create cells that are clinically useful, it is of the utmost importance to devise ways to purify hESC-derived hematopoietic stem cells in order to study them and compare them with their in vivo counterparts.

In this issue of Blood, Vodyanik and colleagues demonstrate that leukosialin (CD43), a protein that is expressed on almost all hematopoietic cells, can be used as a marker to isolate all of the hematopoietic progenitors that are produced by differentiation of hESCs on OP-9, a mouse bone marrow stroma cell line. Previous reports on prospective identification of hESC-derived stem and progenitor hematopoietic cells by flow cytometry were based mostly on the use of the CD34 and CD45 markers. But CD45 is not expressed on the earliest hematopoietic cells, and CD34 is expressed in many lineages, including endothelial and mesenchymal, and therefore lacks specificity to separate the hematopoietic cells from the complex cell mixture that is produced by the coculture of hESCs and OP-9.

Vodyanik and colleagues have now shown that hESC-derived CD34⁺CD43⁺ cells are hematopoietic, while CD34⁺CD43^–CD31⁺ KDR⁺ cells are endothelial and CD34⁺CD43^– CD31^–KDR^– cells are mesenchymal. The authors were also able to distinguish 2 distinct waves of hematopoiesis: after 4 days of coculture, a first wave of CD34⁺CD43⁺ glycophorin A⁺ primitive erythromegakaryocytic progenitors develops. Starting on day 6 of coculture, this first wave is followed by the development of CD34⁺ CD43⁺ glycophorin A^– multipotent lymphohematopoietic progenitors that can give rise in vitro to most myeloid lineages and to B and natural killer (NK) lymphocytes. Whether the hematopoietic cells produced from hESCs can also give rise to T cells, and whether they can be used for long-term reconstitution of lethally irradiated mice, remains to be determined. ▪