Expansion of Cord Blood Stem Cells Mediated by Epigenetic Mechanisms Remain Permissive to External Environmental Cues

Abstract

Widespread use of ex vivo expanded hematopoietic stem cells (HSC) has been largely limited by the lack of proper understanding of factors regulating symmetric self-renewing HSC divisions. We have previously reported that the addition of a hypomethylating agent, 5-aza-2′-deoxyctidine (5azaD) and a histone deacetylase inhibitor, trichostatin A (TSA) in the culture is capable of expanding cord blood (CB) HSC as detected by in vivo SCID repopulating cells (SRC) assay in immunodeficient mice. The increase in SRC during ex vivo expansion culture was associated with greater transcript and protein products of genes implicated in HSC self-renewal (Araki et al. Blood 2007). In order to determine whether variation of exogenous cytokine cocktails added in the culture influences the degree of expansion of HSC treated with 5azaD/TSA, we have cultured CD34+ CB cells in the presence of various cytokine combinations. Interestingly, despite treatment of CB cells with 5azaD/TSA the expansion of stem/progenitor cells varied greatly, depending on the combinations of cytokines used in the culture, ranging between 5 to 12 fold differance. The cytokine combination containing stem cell factor (SCF), Flt3-ligand (FL) and thrombopoietin (TPO) was found to promote maximal expansion of primitive CD34+CD90+ cells following treatment with 5azaD/TSA in comparison to other cytokine combinations used (GM-CSF+SCF+IL-3+IL-6+EPO, SCF+FL+TPO+IL-3, SCF+FL+TPO+IL-6, SCF+FL+TPO+IL-3+IL-6, SCF+IL-3+IL-6). Our results also indicate the importance of sequential addition of 5azaD followed by TSA for the net expansion of HSC. Reversal of the sequence of addition of 5azaD and TSA (TSA followed by 5azaD) resulted in almost complete abrogation of the expansion of primitive CD34+CD90+ cells, and this loss of expansion corresponded with decreased acetylation of histone H4. We have further demonstrated that despite pre-treatment with sequential 5azaD/TSA, various cytokine cocktails in the culture can affect the rate of CD34+CD90+ cell divisions which influences both in vitro clonogenic potential and in vivo SRC potential. The higher in vivo hematopoietic engraftment potential of 5azaD/TSA treated cells in the presence of the optimal cytokine combination (SCF+FL+TPO) is likely due to expansion of a relatively primitive HSC population in the culture which divides slower than the cells expanded in the presence of other cytokine combinations (i.e. SCF+FL+TPO+IL-3+IL-6). Further studies will be needed to understand the molecular mechanism of the loss of functional potential depending on culture conditions. Thus far, in a transwell culture system, CD34+CD90+ cells that have been expanded with 5azaD/TSA show greater migration potential towards stroma derived factor (SDF-1) than CD34+CD90+ cells that have been expanded in cytokines alone without 5azaD/TSA treatment. Most importantly the fraction of migrating cells present in the 5azaD/TSA treated expanded culture was comparable to unmanipulated primary CB CD34+ cells, a likely factor contributing to better engraftment in an immunodeficient mouse model. Our current studies indicate that HSC remain responsive to external humoral influences even after treatment with chromatin modifying agents. The relatively slower cell division rate of CB cells in the presence of 5azaD/TSA might be a critical determinant for the retention of HSC functional capability following ex vivo expansion.