matin) by preventing methylation of DNA and histones by employing 5-azacytidine or tricho- statin A and noticed enhanced self-renewal of murine bone marrow (BM)–derived HSCs that were transplanted into lethally irradiated recipi-

sive—histone marks are physically linked together.1,2 These plastic epigenetic marks prevent undifferentiated ESC differentiation, but in response to environmental cues, they may turn on the specialized transcriptomes to regulate development of more differentiated cells. Although similar epigenetic mechanisms involving “bivalent domains” can occur in adult stem cells,3 the results of genome-wide approaches to better characterize such domains in undifferentiated HSCs are still missing. In this issue of Blood, Chung et al use a genome-wide comparison of DNA methylation to study epigenetic marks in human umbilical cord (UCB) CD34 (undifferentiated) and CD34 (differentiated) cells.4 They found in CD34 cells a presence of characteristic hypomethylation dip around transcription start sites (TSS) of promoters and hypermethylation in flanking regions. The undermethylated DNA methylation pattern near TSS is different in CpG islands (CGI ) and non-CGI (CGI ) genes and seems to be related to an active gene transcription and dynamic chromatin status in a population of primitive HSCs. Furthermore, undifferentiated HSCs exhibited dynamic opentype chromatin associated with transcriptionally active acetylated histones more than terminally differentiated ones. In the next step, the authors inhibited chromatin condensation (heterochromatin) by preventing methylation of DNA and histones by employing 5-azacytidine or trichostatin A and noticed enhanced self-renewal of murine bone marrow (BM)–derived HSCs that were transplanted into lethally irradiated recipients but not of “steady-state” BM-isolated HSCs. Interestingly, similar treatment of more mature cells leads to partial phenotypic dedifferentiation and apoptosis that correlated with the level of their hematopoietic maturation. The authors conclude that the undifferentiated state of hematopoietic cells is characterized by unique epigenetic signature that includes both (1) dynamic chromatin structure and (2) an epigenetic plasticity of gene expression that correlates with the level of differentiation. This paper is important because it addresses several issues related to developmental biology and potential “plasticity” of HSCs. First, the authors clearly demonstrate the dynamic structure of chromatin and its epigenetic modification during differentiation of hematopoietic cells. Second, data presented explain the efficacy of 5-azacytidine and trichostatin A for ex vivo expansion of undifferentiated HSCs at a molecular level. This may be relevant during generation of inducible pluripotent stem cells using the so-called small molecular chromatin modifying agents (eg, BIX-01294 and BayK8644).5 However, because aberrant epigenetic regulation can lead to tumorigenesis and premature stem cell aging,6 full understanding of epigenetic processes that are a kind of “2-edged sword” mechanism is a timely challenge. Finally, additional studies on purified HSCs to elucidate biological significance of unique DNA methylation patterns found in current genomewide analysis will help to better understand epigenetic mechanisms that govern hematopoiesis. Conflict-of-interest disclosure: The authors declare no competing financial interests. ■ REFERENCES