VPA Expanded CD34+CD90+ Cells Despite Having Identical Phenotype Are Functionally Distinct Based on Their Origin

Umbilical cord blood (UCB) as a graft for transplantation in adults often causes prolonged cytopenia, which can be fatal. The anti-convulsant drug valproic acid (VPA) is known to possess histone deacetylase inhibitory activity. We have previously demonstrated that VPA in culture promotes functional hematopoietic stem cell (HSC) maintenance but not expansion (Mahmud et al BBMT 2014). Papa et al. (2018) argued that mitochondrial reprogramming of mature CD34+CD90- cells, rather than replication of immature CD34+CD90+ cells, yields HSC expansion. We asked the question: what fraction of CD34+CD90+ cells in the presence of VPA in culture are derived from relatively mature CD34+CD90- cells vs. immature CD34+CD90+ cells on day 7? A single CD34+ cell yielded 47.48 ± 14.45 CD34+CD90+ cells with VPA in an unsorted CB CD34+ cell culture. Then, CD34+ cells were sorted using flow cytometry into CD90+ and CD90- groups and cultured in the presence of VPA in serum-free media with a cocktail of cytokines. We estimated the contribution of a single CD34+ cell and found that each CD34+ cell gave rise to 64.80 ± 12.42 CD34+CD90+ cells in the CD90+ group and 17.40 ± 9.56 CD34+CD90+ cells in the CD90- group, respectively. The calculated contribution of CD90+ and CD90- groups yielding primitive CD34+CD90+ cells is 28.39 ± 6.75 (*105) compared to the actual count of 41.30 ± 13 (*105) CD34+CD90+ cells. Therefore, the predominant source of CD34+CD90+ cells within CD34+ culture in presence of VPA is derived primarily from replication of CD34+CD90+ cells rather than from relatively mature CD34+CD90- cells. Similarly, VPA treatment on mobilized peripheral blood CD34+ cells displayed similar results. The functional potential of the CD34+CD90+ cells derived from CD90+ vs. CD90- groups was tested. The VPA-treated CD90- group displayed 57.2% less CFU plating efficiency (4.9 ± 2.0%) compared to the CD90+ group (8.56 ± 0.74% CFU). The culture product (VPA+) of the CD90+ and CD90- groups was also transplanted into immunodeficient humanized NSGS mice. NSGS mice bone marrow in the CD90- group showed significantly inferior in vivo transplant potential compared to the CD90+ population-derived culture with VPA after 12 weeks of transplantation (CD90- 1.95 ± 0.83% human blood cell chimerism vs. 14.65 ± 3.16% for the CD90+ group). Notably, our results show that the CD34+CD90+ cells that originate from the relatively mature CD34+CD90- population displays increased mitochondrial mass and mitochondrial membrane potential but exhibit reduced functional potential both in vitro (CFU assays) and in vivo (xenotransplant assays). This suggests that CD34+CD90+ cells that originate from relatively mature CD34+CD90- cells are functionally more mature than CD34+CD90+ cells originating from the expansion of primitive CD34+CD90+ cells after culturing with VPA, despite having identical markers (CD34+CD90+). Specifically, the CD34+CD90+ cells that arise from the relatively mature CD34+CD90- population have increased mitochondrial mass, as detected by mitotracker green using flow cytometry, and increased mitochondrial membrane potential (JC-1 staining) but exhibit notably reduced functional potential both in vitro and in vivo. This suggests that CD34+ cells that originate from CD34+CD90- cells are more mature than CD34+ cells originating from the expansion of CD34+CD90+ cells in culture in the presence of VPA. Taken together, our data show that VPA-expanded CD34+CD90+ cell subpopulation retains higher functional potency in contrast to the minor fraction of CD34+CD90+ cells originating from mature CD34+CD90- cells. These data further emphasize the importance of determining the functional potency of culture-expanded HSC rather than sole reliance on immunophenotypes. Further molecular characterization of CD34+CD90+ cells derived from CD90+ vs. CD90- populations expanded with VPA is currently underway.

No relevant conflicts of interest to declare.