As the progression of biotechnology, it is capable of elucidating the profile of gene expression and their relative abundance in a particular cell type, but this information provides us little about the genes that are not actively transcribed in cells (Spivakov and Fisher, 2007). What is more importantly, it is hard for using expression profiling directly to discriminate between genes that are subject to active repression and those that are not transcribed simply due to the absence or limitation of activating proteins (Spivakov and Fisher, 2007). For this reason, we would not figure out how tissue-specific genes, that will be required for executing later stages in development, are prevented from expression by ES cells, although the potential for their expression is retained (Spivakov and Fisher, 2007). To solve this problem, some underlying mechanisms of gene control in stem cells have been proposed, and they are collectively referred to as epigenetic regulation. These control mechanisms encompass a range of different properties that have been shown to affect gene expression without changes in DNA sequence (Spivakov and Fisher, 2007).
Major epigenetic mechanisms include DNA cytosine methylation, histone modifications such as acetylation and methylation of histone tails, and small non-coding RNA controlled pre- and post-transcriptional regulation of gene expression (Wu and Sun, 2006). Moreover, epigenetic information is known to be able to transmitted through
sequential rounds of cell division (Nakatani et al., 2006; Richards, 2006) because epigenetic marks, including methylated DNA (Jaenisch and Bird, 2003) and modified histones (Henikoff et al., 2004) are propagated at S phase. It is the feature of epigenetic inheritance that lead to the postulation that chromatin has a central role in maintaining transcriptional patterns during development (Spivakov and Fisher, 2007).
B. The roles of chromatin modifications on self-renewal and differentiation of stem cells
Post-translational modification of core histones and methylation of genomic DNA have been revealed to be associated with both chromatin and transcriptional status of genes (Fischle et al., 2003; Hsieh, 2000; Meshorer and Misteli, 2006). During ES-cell differentiation, besides changes in the global genome activity, alterations of histone-modification patterns also occur (Lee et al., 2004). For example, during differentiation, an increase in the silenced chromatin mark tri-methylated residue K9 of histone H3 (H3-triMeK9) and a decrease in the global levels of acetylated histone H3 and H4 (Keohane et al., 1996; Lee et al., 2004; Meshorer et al., 2006), which is usually associated with active chromatin regions, are noted (Meshorer and Misteli, 2006). These findings all suggest that the chromatin of ES-cell is overall either in a more active state or marked with activity-associated histone modifications, and that when differentiation of ES-cell occurs,
the chromatin is transitioned to a transcriptionally less-permissive status (Meshorer and Misteli, 2006). Several repressed heterochromatin marks, such as triMeK9, H3-MeK27, H3-diH3-MeK27, H4-diMeK20 and H4-triMeK20, are found to elevate during RA-induced mouse ES-cell differentiation (Martens et al., 2005). Moreover, the inhibition of mouse ES-cell differentiation after treatment with histone deacetylase (HDAC) inhibitor, trichostatin A (TSA), also indicates the functional relevance to global histone deacetylation during ES-cell differentiation (Lee et al., 2004). Local histone modifications, along with global changes, are believed to be important for the proper control of differentiation-specific genes (Meshorer and Misteli, 2006). Representative example is the promoter of the ES-cell marker, Oct4, which is observed to be enriched for the active mark, H3-triMeK4, in undifferentiated rather than in differentiating ES cells (Lee et al., 2004). Similar findings are also shown for λ5-VpreB1 (Szutorisz et al.,
2005), B-cell differentiation determining gene, and NFM (Kimura et al., 2004), neuronal differentiation determining gene, which all contain active chromatin marks in undifferentiated ES cells despite their inactivity. These observations do indicate that the maintenance of transcriptionally competent chromatin is an active process mediated by histone modifications. The histone modifications help to preserve the pluripotent state of ES cells and mark the transcriptionally competent loci expressed later in the differentiating process (Szutorisz and Dillon, 2005). In addition, there is current evidence
of epigenetic regulation which depicts the existence of the temporary inactivation of differentiation-specific genes in pluripotent cell types (Reik, 2007). These observations unravel that genes required during development and differentiation are those in the homeobox (Hox), distal-less homeobox (Dlx), paired box (Pax) and sine-oculis-related homeobox (Six) gene families. These genes are held repressed in pluripotent ES cells by the Polycomb group (PcG) protein repressive system in mice and humans (Reik, 2007).
PcG protein repressive system marks the histones associated with these genes by inducing methylation of the lysine residue at position 27 of the histone H3 (H3K27) (Azuara et al., 2006; Boyer et al., 2006; Lee et al., 2006). It is also found that when ES cells lose the expression of EED (embryonic ectoderm development), a component of the PcG-protein repressive complex (PRC), the developmental genes are partly derepressed and ES cells are prone to spontaneous differentiation (Azuara et al., 2006; Boyer et al., 2006).
PcG-protein repressive complexes (PRCs) are a subclass of histone modification enzymes that are highly conserved throughout the evolution (Valk-Lingbeek et al., 2004).
Polycomb repressive complex 2 (PRC2) is found to contain both histone deacetylase (HDAC) and histone methyltransferase (HMT) activity, which link hypoacetylation and H3-K9/K27 methylation (Valk-Lingbeek et al., 2004). In another aspect, PRC1 recognizes the H3-K27 methylation mark established by PRC2 through its conserved
chromodomain and takes part in stable maintenance of PRCs mediated gene silencing effect (Wu and Sun, 2006). Chromatin modifying activities opposite to epigenetic control mediated through PRCs also drawed much attention. Testis specific TAF (TBP-associated factor) (TBP-associated trithorax (trx) action (tri-methylation of H3-K4) had been demonstrated to counteract PcG-mediated repression to allow terminal differentiation of Drosophila male germ cell precursors (Chen et al., 2005; Wu and Sun, 2006).
Some developmental genes, however, were present within bivalent chromatin regions which contain both inactivating marks (methylated H3K27) and activating marks (H3K4) (Bernstein et al., 2006; Szutorisz et al., 2005). The bivalent chromatin marks were demonstrated that when the PRCs expressions were downregulated during differentiation and the repressive marks had been removed, these genes were automaticaly poised for transcriptional activation through the H3K4 methylation mark (Reik, 2007). Although epigenetic silencing by PRCs could be mitotically heritable (Ringrose and Paro, 2004), these marks could also be rapidly removed by enzymatic demethylation of H3K27 (Klose et al., 2006). Therefore, in contrast to the terminal silencing achieved by the DNA methylation, developmental genes that were silenced by PRCs in pluripotent tissues required repressive marks to be rapidly and flexibly moved when differentiation begain (Reik, 2007). Another group of genes which encoded pluripotency-sustaining transcription factors, such as Oct4 and Nanog, were required for
early development or for germ-cell development only. These pluripotency-associated genes expressed by ES cells but silent during the differentiation of these cells were also known to be mediated by epigenetic regulation with a defined kinetics of acquiring repressive histone modifications and DNA methylation (Feldman et al., 2006; Reik, 2007).
Albeit silence of genetic element could be achieved through histone modifications, these easily reversible modifications were not good gatekeeper for long-term silence (Shi et al., 2004; Takeuchi et al., 2006). Thus, prolonged silence of genetic element must be mediated by an additional epigenetic mechanism. An important component of this process is DNA methylation (Miranda and Jones, 2007). It was noticed that within gene promoters, even when the repressive marks were removed, DNA methylation still prevented the reactivation of silent genes (McGarvey et al., 2007). It is also found that DNA methylation was important for many cellular processes including the silence of repetitive elements, X-inactivation, imprinting and development. The roles of DNA methylation in these processes ensured the daughter cells to retain the same expression pattern as the precursor cells (Miranda and Jones, 2007). DNA methylation is a covalent modification in which the 5’ position of cytosine is methylated in a reaction catalyzed by DNA methyltransferases (DNMTs) with S-adenosyl-methionine as the methyl donor (Miranda and Jones, 2007). In mammals, methylation of cytosines in cytosine guanine
dinucleotide (CpG) island has been known to play a crucial role in mediating epigenetic gene silencing through two possible mechanisms. First, it has been proved that cytosine methylation can directly silence gene expression by inhibiting DNA binding factors to its recognition gene area (Watt and Molloy, 1988). Second, some authors (Boyes and Bird, 1991) and other investigators also prove that methylated-CpG can recruit Methyl-CpG-binding proteins (MBPs), which in turn function to silence transcription and modify the surrounding chromatin structure. In mammals, four DNA methyltransferases (DNMTs) have been identified. Among them, DNMT1 has been known to maintain the pre-existing methylation state of genome during DNA replication (Leonhardt et al., 1992). DNMT3a and DNMT3b are de novo methyltransferases which will target unmethylated CpG sites (Okano et al., 1999). DNMT2, which has been proved to have weak methyltransferase activity in vitro (Hermann et al., 2003), is responsible to methylate tRNA (Goll et al., 2006).
In differentiating ES cells, increase of CpG island DNA methylation with enhanced expression of DNMTs was also noticed (Kremenskoy et al., 2003; Shen et al., 2006), and the deletion of three major DNMTs would cause hypomethylation and thorough blockage of differentiation of ES cells (Carlone et al., 2005; Jackson et al., 2004). These findings plus the fact that global methylation marks are erased during early embryogenesis and then increase during in vitro expansion (Maitra et al., 2005) suggest that differential
changes in CpG island DNA methylation profile may serve as an indicator of
“primitiveness” or un-commiment status of stem cells. What’s noteworthy, several studies (Hattori et al., 2004; Taylor and Jones, 1979; Tsuji-Takayama et al., 2004) even observed the phenomenon of dedifferentiation caused by treating cells with demethylation agent 5-azacytidine (5-AzaC), which implied that DNA demethylation might cause a reversion of cells to a more pluripotent state (Meshorer and Misteli, 2006).
In 1979, Taylor and Jones (Taylor and Jones, 1979) firstly described the phenomenon that by treating C3H/10T1/2 cells with the demethylating agent 5-azacytidine (5-AzaC), differentiate into striated muscle cells, adipocytes and chondrocytes was found to be induced. The authors explained that the conversion of these cells to new phenotypes was caused by a reversion of cells toward a primitive pluripotent state and subsequently gave rise to other lineages. In consistent with this viewpoint, both partially differentiated ES cells (Tsuji-Takayama et al., 2004) and trophoblast stem cells (Hattori et al., 2004) treated with 5-AzaC were found to be induced toward trend of dedifferentiation as well.