Distinct Mechanisms Underlying the Ex Vivo Expansion of Human Cord Blood Stem Cells with Different Strategies Currently Used for Allogeneic Transplantation

Umbilical cord blood (UCB) is an alternative source of hematopoietic stem cells (HSCs) for patients that require allogeneic stem cell transplantation. The use of UCB in adults is restricted due to the limited number of HSCs within a single unit. This limitation can be overcome by ex-vivo expansion.

Several strategies have successfully expanded the numbers of clinically functional HSCs. Such strategies, including the aryl hydrocarbon receptor antagonist, SR1, pyrimidoindole derivative, UM171, nicotinamide, NAM and more recently, valproic acid (VPA), have entered clinical development. HSC grafts expanded by SR1 and UM171 have been reported to establish rapid and sustained hematological reconstitution in allogeneic transplant recipients [1-3]. Our group has started a clinical trial using VPA-expanded allogeneic UCB grafts (NCT03885947). Initial data have shown multi-lineage donor cell reconstitution with especially rapid T cell and platelet engraftment in patients with treatment refractory acute leukemias.

To gain insight into the mechanisms underlying these clinical HSC expansion procedures, we recreated the different approaches to compare the various expanded grafts. Although there is successful clinical data with the NAM approach, we were unable to recreate these conditions in our laboratory and therefore, we could not compare these grafts to those expanded with SR1, UM171 and VPA [4-7].

These comparative studies indicated that each strategy increased to varying degrees both the percentage and absolute numbers of CD34⁺CD133⁺CD90⁺EPCR⁺ as opposed to cultures treated with their respective cytokine cocktail alone. Although SR1 treatment resulted in the generation of greater numbers of total nucleated cells from purified 1x10⁶ UCB-CD34⁺ cells (16x10⁷ in SR1 vs 8x10⁷ in UM171 and 7x10⁷ in VPA), the numbers of CD34⁺CD133⁺CD90⁺EPCR⁺ cells (4x10⁵ in SR1 vs, 7x10⁵ in UM171 and 8.7x10⁵ in VPA) as well as of other CD34⁺ subpopulations with different degrees of stem cell activity were lower than those observed in VPA and UM171-expanded cultures.

In order to understand the underlying mechanism by which these agents act, we isolated CD34⁺CD90^-EPCR^- cells and treated with VPA, SR1 or UM171. VPA treatment for 24 hrs resulted in acquisition of CD90 and EPCR expression (80% and 25% of cells acquired CD90 and EPCR, respectively). UM171 treatment led to a modest increase in cells expressing EPCR and CD90. By contrast, SR1 treatment did not induce a phenotypic switch.

Comparison of RNA-seq data indicated a strong induction of CD90, EPCR and CD133 in the CD34⁺ grafts expanded with VPA and UM171, but not in grafts generated with SR1. Twenty five out of 53 and 5 out of 9 genes upregulated more than 2 fold by UM171 and SR1, respectively, were also upregulated in VPA-expanded grafts. Eight out of 39 and 6 out of 77 genes downregulated more than 1.5 fold by UM171 and SR1, respectively, were also downregulated in VPA-expanded grafts. However, grafts expanded with each of these agents shared only three upregulated and three downregulated genes suggesting that different mechanisms underlie the actions of each of these agents. In fact, enhanced histone acetylation induced by VPA was not observed in SR1 and UM171 expanded cells. Both UM171 and VPA did not signal through the AHR pathway, which is required for SR1-mediated expansion. These data suggest that these agents target distinct pathways leading to ex vivo HSC expansion.

Since mitochondrial activity is an additional critical regulator of the HSC fate, we examined the mitochondrial profile. VPA-expanded HSCs exhibited a remodeled mitochondrial network associated with low levels of mitochondrial ROS and mass. HSCs expanded with UM171 unlike SR1 displayed low ROS levels. Neither UM171 nor SR1 impacted mitochondrial mass.

In vivo transplantation of UM171 and VPA expanded grafts in NSG mice showed similar degrees of human cell chimerism (~25%). Whereas both expanded grafts were capable of establishing multilineage hematopoiesis, we found a greater degree of CD34⁺ and T cell chimerism in NSG mice engrafted with VPA-expanded grafts. Studies of the engraftment potential of SR1 expanded grafts are ongoing.

In conclusion, our findings suggest that although VPA, UM171 and SR1 expand greater numbers of human HSCs that are useful for allogeneic stem cell transplantation, their underlying mechanisms of action and pathways involved are different.