Genetic control of hematopoietic stem cell differentiation
Differentiation of stem cells is associated with two fundamental processes, reduction in self-renewal potential and step-wise acquisition of a specific lineage identity. These reciprocal processes are controlled by competing genetic programs. If a stem cell is triggered to begin differentiating, genes that maintain self-renewal are switched off, whereas genes that enforce differentiation are switched on. The formation of early hematopoietic progenitors from stem cells is orchestrated by a relatively small number of transcription factors. Among them are PU.1, CCAAT/enhancer binding protein alpha (C/EBPalpha), growth factor independent 1 (GFI1), interferon-regulatory factor 8 (IRF8), Runt-related transcription factor 1 (RUNX1) and stem-cell leukemia factor (SCL). Mice in which these genes have been knocked out displayed profound hematopoietic defects. Moreover, these transcription factors were shown to regulate a broad range of pivotal target genes, thereby directly programming precursors to differentiate along a complex developmental pathway. A block in normal differentiation is a major contributing factor towards the development of solid tumors and leukemias and cells from leukemia patients frequently harbor mutated or dysregulated transcription factor genes. This suggests that altered transcription factor activity is a major driving force behind the pathology of transformation and the development of cancer stem cells.

One of the main interests of our laboratory is to understand how transcription factors direct normal stem cell functions, such as self-renewal and differentiation, how they program precursors to adopt a certain lineage choice and how disruption of transcription factor activity leads to cancer (stem) cell transformation. Using both transgenic and knockout mouse models, we are particularly interested in discovering crucial molecular up- and downstream mechanisms that regulate the expression and function of transcription factors. A current research focus in our laboratory is on PU.1. The Ets-family member PU.1 is essential for both myeloid and lymphoid lineages. PU.1 knockout mice exhibit early lethality and lack of B-lymphocytes and mature myeloid cells in fetal livers. In addition, PU.1 is important for HSC self-renewal and differentiation into the earliest myeloid and lymphoid progenitors. Furthermore, PU.1 must be properly downregulated in early thymocytes to allow normal T cell development, since enforced PU.1 expression in thymic organ cultures completely blocked T cell production. We and others could reveal that graded changes in PU.1 concentrations have drastic effects on lineage fate decisions. Therefore, a greater understanding of PU.1 gene regulation is the key to deciphering its role in normal hematopoiesis and malignant transformation.

Figure 1: Regulated interaction between cis-regulatory elements orchestrates the PU.1 expression pattern (Leddin et al. Blood 2011).

Although the well-controlled tissue-specific expression of PU.1 is essential for hematopoiesis and leukemia, little is known of how this pattern is established. We identified an upstream regulatory cis-element (URE) whose targeted deletion in mice decreases PU.1 expression and causes leukemia. However, we showed further that the URE alone is insufficient to confer physiological PU.1 expression, but requires the cooperation with other, previously unidentified elements. Using a combination of transgenic studies, global chromatin assays and detailed molecular analyses we presented evidence that PU.1 is regulated by a novel mechanism involving cross-talk between different cis-elements together with lineage-restricted autoregulation. In this model, PU.1 regulates its expression in B cells and macrophages by differentially associating with cell-type specific transcription factors at one of its cis-regulatory elements to establish differential activity patterns at other elements (Figure 1).

Epigenetic control of hematopoietic stem cell function
Regulation of the diverse functional repertoire of HSCs requires the coordinated action of transcription factors. However, the activity of most transcription factors critically relies on the recruitment of cofactors, many of which control gene expression by catalyzing epigenetic modifications of chromatin. However, the functional impact of epigenetic modification mechanisms on coordination of stem cell fate programs is still poorly understood.

Figure 2: Model of DNA methylation dosage effects on stem cell multipotency (Bröske et al. Nature Genetics 2009).

Methylation of CpG dinucleotides within the DNA is a major epigenetic modification, which in mammals is controlled by at least 3 different DNA-methyltranferases (DNMTs): DNMT3a and -b for de novo methylation and DNMT1 for methylation maintenance. The impact of methylation on stem cell features has been studied in embryonic stem (ES) cells, but little is known about its function in somatic stem cells in vivo. Recent advances in the genome-wide mapping of DNA methylation demonstrated that methylation of CpGs are dynamic epigenetic marks that undergo extensive changes during cellular differentiation. However, whether and how these changes are required for cell fate choice, in particular for that of stem cells, remains unknown. Moreover, altered DNA methylation is a hallmark of cancer, and drugs targeting methylating enzymes are used in cancer therapy. The relationship between tumor-associated alterations in methylation and cancer stem cell properties is still elusive. For that reason, it is of major general interest to understand whether and how methylation mechanisms control key features of normal and transformed stem cells.

We could show that alternative functional programs of HSCs are governed by gradual differences in the methylation level (Figure 2). Constitutive methylation is essential for HSC self-renewal, but dispensable for homing, cell cycle control and suppression of apoptosis. Remarkably, HSCs from mice with reduced DNMT 1 activity fail to suppress key myeloerythroid regulators and as a consequence can differentiate into myeloerythroid but not into lymphoid progeny. We revealed that a similar methylation dosage effect controls stem cell function in leukemia. Thus, our data identified DNA methylation as an essential epigenetic mechanism to protect stem cells from premature activation of predominant differentiation programs and suggest that methylation dynamics determines stem cell functions in tissue homeostasis and cancer. Consequently, these results provide the hope that demethylating drugs may be instrumental to impair the function of cancer stem cells in cancer therapy.