NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Brown TA. Genomes. 2nd edition. Oxford: Wiley-Liss; 2002.

Cover of Genomes

Genomes. 2nd edition.

Show details

Chapter 8Accessing the Genome

Learning outcomes

When you have read Chapter 8, you should be able to:

  • Explain how chromatin structure influences genome expression
  • Describe the internal architecture of the eukaryotic nucleus
  • Distinguish between the terms ‘constitutive heterochromatin’, ‘facultative heterochromatin’ and ‘euchromatin’
  • Discuss the key features of functional domains, insulators, and locus control regions, and describe the experimental evidence supporting our current knowledge of these structures
  • Describe the various types of chemical modification that can be made to histone proteins, and link this information to the concept of the ‘histone code’
  • State why nucleosome positioning is important in gene expression and give details of a protein complex involved in nucleosome remodeling
  • Explain how DNA methylation is carried out and describe the importance of methylation in silencing the genome
  • Give details of the involvement of DNA methylation in genomic imprinting and X inactivation

When one looks at a genome sequence written out as a series of As, Cs, Gs and Ts, or drawn as a map with the genes indicated by boxes on a string of DNA (as in Figure 1.14, for example), there is a tendency to imagine that all parts of the genome are readily accessible to the DNA-binding proteins that are responsible for its expression. In reality, the situation is very different. The DNA in the nucleus of a eukaryotic cell or the nucleoid of a prokaryote is attached to a variety of proteins that are not directly involved in genome expression and which must be displaced in order for the RNA polymerase and other expression proteins to gain access to the genes. We know very little about these events in prokaryotes, a reflection of our generally poor knowledge about the physical organization of the prokaryotic genome (Section 2.3.1), but we are beginning to understand how the packaging of DNA into chromatin (Section 2.2.1) influences genome expression in eukaryotes. This is an exciting area of molecular biology, with recent research indicating that histones and other packaging proteins are not simply inert structures around which the DNA is wound, but instead are active participants in the processes that determine which parts of the genome are expressed in an individual cell. Many of the discoveries in this area have been driven by new insights into the substructure of the nucleus, and it is with this topic that we begin the chapter.

8.1. Inside the Nucleus

The light microscopy and early electron microscopy studies of eukaryotic cells revealed very few internal features within the nucleus. This apparent lack of structure led to the view that the nucleus has a relatively homogeneous architecture, a typical ‘black box’ in common parlance. In recent years this interpretation has been overthrown and we now appreciate that the nucleus has a complex internal structure that is related to the variety of biochemical activities that it must carry out. Indeed, the inside of the nucleus is just as complex as the cytoplasm of the cell, the only difference being that, in contrast to the cytoplasm, the functional compartments within the nucleus are not individually enclosed by membranes, and so are not visible when the cell is observed using conventional light or electron microscopy techniques.

8.1.1. The internal architecture of the eukaryotic nucleus

The revised picture of nuclear structure has emerged from two novel types of microscopy analysis. First, conventional electron microscopy has been supplemented by examination of mammalian cells that have been prepared in a special way. After dissolution of membranes by soaking in a mild non-ionic detergent such as one of the Tween compounds, followed by treatment with a deoxyribonuclease to degrade the nuclear DNA, and salt extraction to remove the chemically basic histone proteins, the nuclear substructure has been revealed as a complex network of protein and RNA fibrils, called the nuclear matrix (Figure 8.1A). The matrix permeates the entire nucleus and includes regions defined as the chromosome scaffold, which changes its structure during cell division, resulting in condensation of the chromosomes into their metaphase forms (see Figure 2.7).

Figure 8.1. The internal architecture of the eukaryotic nucleus.

Figure 8.1

The internal architecture of the eukaryotic nucleus. (A) Transmission electron micrograph showing the nuclear matrix of a cultured human HeLa cell. Cells were treated with a non-ionic detergent to remove membranes, digested with a deoxyribonuclease to (more...)

A second novel type of microscopy has involved the use of fluorescent labeling, designed specifically to reveal areas within the nucleus where particular biochemical activities are occurring. The nucleolus (Figure 8.1B), which is the center for synthesis and processing of rRNA molecules, had been recognized for many years as it is the one structure within the nucleus that can be seen by conventional electron microscopy. Fluorescent labeling directed at the proteins involved in RNA splicing (Sections 1.2.1 and 10.1.3) has shown that this activity is also localized into distinct regions (Figure 8.1C), although these are more widely distributed and less well defined than the nucleoli. Other structures, such as Cajal bodies (visible in Figure 8.1B), whose functions are not yet understood, are also seen after fluorescent labeling (Lewis and Tollervey, 2000).

The complexity of the nuclear matrix, as shown by Figure 8.1A, could be taken as an indication that the nucleus has a static internal environment, with limited movement of molecules from one site to another. Another new microscopy technique, called fluorescence recovery after photobleaching (FRAP; Technical Note 8.1), which enables the movement of proteins within the nucleus to be visualized, shows that this is not the case. The migration of nuclear proteins does not occur as rapidly as would be expected if their movement were totally unhindered, which is entirely expected in view of the large amounts of DNA and RNA in the nucleus, but it is still possible for a protein to traverse the entire diameter of a nucleus in a matter of minutes (Misteli, 2001). Proteins involved in genome expression therefore have the freedom needed to move from one activity site to another, as dictated by the changing requirements of the cell. In particular, the linker histones (Section 2.2.1) continually detach and reattach to their binding sites on the genome (Lever et al., 2000; Misteli et al., 2000). This discovery is important because it emphasizes that the DNA–protein complexes that make up chromatin are dynamic, an observation that has considerable relevance to genome expression, as we will discover in the next section.

Box Icon

Box 8.1

Fluorescence recovery after photobleaching (FRAP). Visualization of protein mobility in living nuclei. FRAP is perhaps the most informative of the various innovative microscopy techniques that have opened up our understanding of nuclear substructure. (more...)

Box Icon

Box 8.1

Accessing the prokaryotic genome. Chapter 8 is exclusively about eukaryotic genomes. We simply do not know enough about the physical structures of bacterial and archaeal genomes to be able to state, with any degree of confidence, the extent to which expression (more...)

8.1.2. Chromatin domains

In Section 2.2.1 we learnt that chromatin is the complex of genomic DNA and chromosomal proteins present in the eukaryotic nucleus. Chromatin structure is hierarchic, ranging from the two lowest levels of DNA packaging – the nucleosome and the 30 nm chromatin fiber (see Figures 2.5 and 2.6) – to the metaphase chromosomes, which represent the most compact form of chromatin in eukaryotes and occur only during nuclear division. After division, the chromosomes become less compact and cannot be distinguished as individual structures. When non-dividing nuclei are examined by light microscopy all that can be seen is a mixture of lightly and darkly staining areas within the nucleus. The dark areas, which tend to be concentrated around the periphery of the nucleus, are called heterochromatin and contain DNA that is still in a relatively compact organization, although still less compact than in the metaphase structure. Two types of heterochromatin are recognized:

  • Constitutive heterochromatin is a permanent feature of all cells and represents DNA that contains no genes and so can always be retained in a compact organization. This fraction includes centromeric and telomeric DNA as well as certain regions of some other chromosomes. For example, most of the human Y chromosome is made of constitutive heterochromatin (see Figure 2.8).
  • Facultative heterochromatin is not a permanent feature but is seen in some cells some of the time. Facultative heterochromatin is thought to contain genes that are inactive in some cells or at some periods of the cell cycle. When these genes are inactive, their DNA regions are compacted into heterochromatin.
    It is assumed that the organization of heterochromatin is so compact that proteins involved in gene expression simply cannot access the DNA. In contrast, the remaining regions of chromosomal DNA, the parts that contain active genes, are less compact and permit entry of the expression proteins. These regions are called euchromatin and they are dispersed throughout the nucleus. The exact organization of the DNA within euchromatin is not known, but with the electron microscope it is possible to see loops of DNA within the euchromatin regions, each loop between 40 and 100 kb in length and predominantly in the form of the 30 nm chromatin fiber. The loops are attached to the nuclear matrix via AT-rich DNA segments called matrix-associated regions (MARs) or scaffold attachment regions (SARs) (Figure 8.2).
    The loops of DNA between the nuclear matrix attachment points are called structural domains. An intriguing question is the precise relationship between these and the functional domains that can be discerned when the region of DNA around an expressed gene or set of genes is examined. A functional domain is delineated by treating a region of purified chromatin with deoxyribonuclease I (DNase I) which, being a DNA-binding protein, cannot gain access to the more compacted regions of DNA (Figure 8.3). Regions sensitive to DNase I extend to either side of a gene or set of genes that is being expressed, indicating that in this area the chromatin has a more open organization, although it is not clear whether this organization is the 30 nm fiber or the ‘beads-on-a-string’ structure (see Figure 2.5A). Is there a correspondence between structural and functional domains? Intuition suggests that there should be, and some MARs, which mark the limits of a structural domain, are also located at the boundary of a functional domain. But the correspondence does not seem to be complete because some structural domains contain genes that are not functionally related, and the boundaries of some structural domains lie within genes (Wolffe, 1995).

Figure 8.2. A scheme for organization of DNA in the nucleus.

Figure 8.2

A scheme for organization of DNA in the nucleus. The nuclear matrix is a fibrous protein-based structure whose precise composition and arrangement in the nucleus has not been described. Euchromatin, predominantly in the form of the 30 nm chromatin (more...)

Figure 8.3. A functional domain in a DNase I sensitive region.

Figure 8.3

A functional domain in a DNase I sensitive region.

Functional domains are defined by insulators

The boundaries of functional domains are marked by sequences, 1–2 kb in length, called insulators (Bell et al., 2001). Insulator sequences were first discovered in Drosophila and have now been identified in a range of eukaryotes. The best studied are the pair of sequences called scs and scs′ (scs stands for ‘specialized chromatin structure’), which are located either side of the two hsp70 genes in the fruit-fly genome (Figure 8.4).

Figure 8.4. Insulator sequences in the fruit-fly genome.

Figure 8.4

Insulator sequences in the fruit-fly genome. The diagram shows the region of the Drosophila genome containing the two hsp70 genes. The insulator sequences scs and scs′ are either side of the gene pair. The arrows below the two genes indicate that (more...)

Insulators display two special properties related to their role as the delimiters of functional domains. The first is their ability to overcome the positional effect that occurs during a gene cloning experiment with a eukaryotic host. The positional effect refers to the variability in gene expression that occurs after a new gene has been inserted into a eukaryotic chromosome. It is thought to result from the random nature of the insertion event, which could deliver the gene to a region of highly packaged chromatin, where it will be inactive, or into an area of open chromatin, where it will be expressed (Figure 8.5A). The ability of scs and scs′ to overcome the positional effect was demonstrated by placing them either side of a fruit-fly gene for eye color (Kellum and Schedl, 1991). When flanked by the insulators, this gene was always highly expressed when inserted back into the Drosophila genome, in contrast to the variable expression that was seen when the gene was cloned without the insulators (Figure 8.5B). The deduction from this and related experiments is that insulators can bring about modifications to chromatin packaging and hence establish a functional domain when inserted into a new site in the genome.

Figure 8.5. The positional effect.

Figure 8.5

The positional effect. (A) A cloned gene that is inserted into a region of highly packaged chromatin will be inactive, but one inserted into open chromatin will be expressed. (B) The results of cloning experiments without (red) and with (blue) insulator (more...)

Insulators also maintain the independence of each functional domain, preventing ‘cross-talk’ between adjacent domains. If scs or scs′ is excised from its normal location and re-inserted between a gene and the upstream regulatory modules that control expression of that gene, then the gene no longer responds to its regulatory modules: it becomes ‘insulated’ from their effects (Figure 8.6A). This observation suggests that, in their normal positions, insulators prevent the genes within a domain from being influenced by the regulatory modules present in an adjacent domain (Figure 8.6B).

Figure 8.6. Insulators maintain the independence of a functional domain.

Figure 8.6

Insulators maintain the independence of a functional domain. (A) When placed between a gene and its upstream regulatory modules, an insulator sequence prevents the regulatory signals from reaching the gene. (B) In their normal positions, insulators prevent (more...)

How insulators carry out their roles is not yet known but it is presumed that the functional component is not the insulating sequence itself but the DNA-binding proteins, such as Su(Hw) in Drosophila, that attach specifically to insulators. As well as binding to insulators, these proteins form associations with the nuclear matrix (Gerasimova et al., 2000), possibly indicating that the functional domains that they define are also structural domains within the chromatin. This is an attractive hypothesis that can be tied in with the ability of insulators to establish open chromatin regions and to prevent cross-talk between functional domains, but it implies that insulators contain MAR sequences, which has not been proven. An equivalence between functional and structural domains therefore remains elusive.

Some functional domains contain locus control regions

The formation and maintenance of an open functional domain, at least for some domains, is the job of a DNA sequence called the locus control region or LCR (Li et al., 1999). Like insulators, an LCR can overcome the positional effect when linked to a new gene that is inserted into a eukaryotic chromosome. Unlike insulators, an LCR also stimulates the expression of genes contained within its functional domain.

LCRs were first discovered during a study of the human β-globin genes (Section 2.2.1) and are now thought to be involved in expression of many genes that are active in only some tissues or during certain developmental stages. The globin LCR is contained in a stretch of DNA some 12 kb in length, positioned upstream of the genes in the 60-kb β-globin functional domain (Figure 8.7). The LCR was initially identified during studies of individuals with thalassemia, a blood disease that results from defects in the α- or β-globin proteins. Many thalassemias result from mutations in the coding regions of the globin genes, but a few were shown to map to a 12-kb region upstream of the β-globin gene cluster, the region now called the LCR. The ability of mutations in the LCR to cause thalassemia is a clear indication that disruption of the LCR results in a loss of globin gene expression.

Figure 8.7. DNase I hypersensitive sites indicate the position of the locus control region for the human β-globin gene cluster.

Figure 8.7

DNase I hypersensitive sites indicate the position of the locus control region for the human β-globin gene cluster. A series of hypersensitive sites are located in the 20 kb of DNA upstream of the start of the β-globin gene cluster. (more...)

More detailed study of the β-globin LCR has shown that it contains five separate DNase I hypersensitive sites, short regions of DNA that are cleaved by DNase I more easily than other parts of the functional domain. These sites are thought to coincide with positions where nucleosomes have been modified or are absent and which are therefore accessible to binding proteins that attach to the DNA. It is these proteins, not the DNA sequence of the LCR, that control the chromatin structure within the functional domain. Exactly how, and in response to what biochemical signals, is not known.

DNase I hypersensitive sites also occur immediately upstream of each of the genes in the β-globin LCR, at the positions where the transcription initiation complex is assembled on the DNA (Section 9.2.3). These assembly positions illustrate an interesting feature of hypersensitive sites: they are not invariant components of a functional domain. Recall that the different β-type globin genes are expressed at different stages of the human developmental cycle, ε being active in the early embryo, Gγ and Aγ in the fetus, and δ and β in the adult (see Figure 2.14). Only when the gene is active is its assembly position for the transcription initiation complex marked by a hypersensitive site. Initially it was thought that this was an effect of the differential expression of these genes, in other words that in the absence of gene activity it was possible for nucleosomes to cover the assembly site, presumably to be pushed to one side when it became time to express the gene. Now it is thought that the presence or absence of nucleosomes is a cause of gene expression, the gene being switched off if nucleosomes cover the assembly site, or switched on if access to the site is open.

8.2. Chromatin Modifications and Genome Expression

The previous sections have introduced us to two ways in which chromatin structure can influence genome expression (Figure 8.8). First, the degree of chromatin packaging displayed by a segment of a chromosome determines whether or not genes within that segment are expressed. Second, if a gene is accessible, then its transcription is influenced by the precise nature and positioning of the nucleosomes in the region where the transcription initiation complex will be assembled. Significant advances in understanding both types of chromatin modification have been made in recent years, and we now recognize that specific regions of the genome can be either activated or silenced by processes that involve modification of chromatin structure. We know rather more about the activation processes, so we will begin with these.

Figure 8.8. Two ways in which chromatin structure can influence gene expression.

Figure 8.8

Two ways in which chromatin structure can influence gene expression. A region of unpackaged chromatin in which the genes are accessible is flanked by two more compact segments. Within the unpackaged region, the positioning of the nucleosomes influences (more...)

8.2.1. Activating the genome

Nucleosomes appear to be the primary determinants of genome activity in eukaryotes, not only by virtue of their positioning on a strand of DNA, but also because the precise chemical structure of the histone proteins contained within nucleosomes is the major factor determining the degree of packaging displayed by a segment of chromatin.

Histone modifications determine chromatin structure

Histone proteins can undergo various types of modification, the best studied of these being histone acetylation – the attachment of acetyl groups to lysine amino acids in the N-terminal regions of each of the core molecules. These N termini form tails that protrude from the nucleosome core octamer (Figure 8.9) and their acetylation reduces the affinity of the histones for DNA and possibly also reduces the interaction between individual nucleosomes that leads to formation of the 30 nm chromatin fiber. The histones in heterochromatin are generally unacetylated whereas those in functional domains are acetylated, a clear indication that this type of modification is linked to DNA packaging.

Figure 8.9. Two views of the nucleosome core octamer.

Figure 8.9

Two views of the nucleosome core octamer. The view on the left is downwards from the top of the barrel-shaped octamer; the view on the right is from the side. The two strands of the DNA double helix wrapped around the octamer are shown in brown and green. (more...)

The relevance of histone acetylation to genome expression was underlined in 1996 when, after several years of trying, the first examples of histone acetyltransferases (HATs) – the enzymes that add acetyl groups to histones – were identified (Pennisi, 1997). It was realized that some proteins that had already been shown to have important influences on genome expression had HAT activity. For example, one of the first HATs to be discovered, the Tetrahymena protein called p55, was shown to be a homolog of a yeast protein, GCN5, which was known to activate assembly of the transcription initiation complex (Brownell et al., 1996; Section 9.3.2). Similarly, the mammalian protein called p300/CBP, which had been ascribed a clearly defined role in activation of a variety of genes, was found to be a HAT (Bannister and Kouzarides, 1996). These observations, plus the demonstration that different types of cell display different patterns of histone acetylation, are clear indications that histone acetylation plays a prominent role in regulating genome expression.

Individual HATs can acetylate histones in the test tube but have negligible activity on intact nucleosomes, indicating that, in the nucleus, HATs almost certainly do not work singly, but instead form multiprotein complexes, such as the ADA and SAGA complexes of yeast and the TFTC complex of humans. Different complexes appear to acetylate different histones and some can also acetylate other proteins involved in genome expression, such as the general transcription factors TFIIE and TFIIF, which we will meet in Section 9.2.3. There are also indications that in addition to local modifications to histone proteins in the regions surrounding expressed genes, HATs can also carry out more general modifications on a global scale throughout the entire genome (Berger, 2000).

Acetylation is not the only type of histone modification. The tails of the core histones also have attachment sites for methyl and phosphate groups and for the common (‘ubiquitous’) protein called ubiquitin. Although information on these modifications is limited, it is clear that they too can influence chromatin structure and have a significant impact on cellular activity. For example, phosphorylation of histone H3 and of the linker histone has been associated with formation of metaphase chromosomes (Bradbury, 1992), and ubiquitination of histone H2B is part of the general role that ubiquitin plays in control of the cell cycle (Robzyk et al., 2000). The effects of methylation of a pair of lysine amino acids at the fourth and ninth positions from the N-terminus of histone H3 are particularly interesting. Methylation of lysine-9 forms a binding site for the HP1 protein which induces chromatin packaging and silences gene expression (Lachner et al., 2001; Bannister et al., 2001), but methylation of lysine-4 has the opposite effect and promotes an open chromatin structure. Within the β-globin functional domain, and probably elsewhere, lysine-4 methylation is closely correlated with acetylation of histone H3 (Litt et al., 2001), and the two types of modification may work hand in hand to activate regions of chromatin. Our growing awareness of the variety of histone modifications that occur, and of the way in which different modifications work together, has led to the suggestion that there is a histone code, by which the pattern of chemical modifications specifies which regions of the genome are expressed at a particular time (Strahl and Allis, 2000; Jenuwein and Allis, 2001).

Nucleosome remodeling influences the expression of individual genes

The second type of chromatin modification that can influence genome expression is nucleosome remodeling. This term refers to the modification or repositioning of nucleosomes within a short region of the genome, so that DNA-binding proteins can gain access to their attachment sites. Unlike acetylation and the other chemical modifications described in the previous section, nucleosome remodeling does not involve covalent alterations to histone molecules. Instead, remodeling is induced by an energy-dependent process that weakens the contact between the nucleosome and the DNA with which it is associated. Three distinct types of change can occur (Figure 8.10):

Figure 8.10. Nucleosome remodeling, sliding and transfer.

Figure 8.10

Nucleosome remodeling, sliding and transfer.

  • Remodeling, in the strict sense, involves a change in the structure of the nucleosome, but no change in its position. The nature of the structural change is not known, but when induced in vitro the outcome is a doubling in size of the nucleosome and an increased DNase sensitivity of the attached DNA.
  • Sliding, or cis -displacement, physically moves the nucleosome along the DNA.
  • Transfer, or trans -displacement, results in the nucleosome being transferred to a second DNA molecule, or to a non-adjacent part of the same molecule.
    As with HATs, the proteins responsible for nucleosome remodeling work together in large complexes. One of these is Swi/Snf, made up of at least 11 proteins, which is present in many eukaryotes (Sudarsanam and Winston, 2000). Little is currently known about the way in which Swi/Snf, or any other nucleosome remodeling complex, carries out is role in increasing access to the genome. None of the components of Swi/Snf appears to have a DNA-binding capability, so the complex must be recruited to its target site by additional proteins. Interactions have been detected between Swi/Snf and HATs (Syntichaki et al., 2000), suggesting that nucleosome remodeling might occur in conjunction with histone acetylation. This is an attractive hypothesis because it links the two activities that are currently looked on as central to genome activation. But there are problems because Swi/Snf does not appear to have a global effect across an entire genome, but instead influences expression at only a limited number of positions; in the case of yeast, perhaps at no more than 6% of all the genes in the genome. This observation suggests that the more important interactions might be between Swi/Snf and proteins that target a limited set of genes, rather than the HATs which work throughout the genome. The most likely candidates are transcription activators (Section 9.3.2), each of which is specific for a limited set of genes and some of which form associations with Swi/Snf in vitro.

8.2.2. Silencing the genome

If mechanisms exist for activating parts of the genome then logic dictates that there must also be complementary processes that silence regions whose expression products are not needed. Silencing of specific genes can be achieved simply by reversing the activating effects of histone acetylation, but acetylation–deacetylation is not the only process that influences genome expression. Direct methylation of DNA also has a silencing effect, and special processes exist for inactivating individual chromosomes. We will look at each of these silencing mechanisms in turn.

Histone deacetylation is one way of repressing gene expression

One way in which silencing can be implemented is by removing acetyl groups from histone tails, and hence reversing the transcription-activating effects of the HATs described above. This is the role of the histone deacetylases (HDACs). The link between HDAC activity and gene silencing was established in 1996, when mammalian HDAC1, the first of these enzymes to be discovered, was shown to be related to the yeast protein called Rpd3, which was known to be a repressor of transcription (Taunton et al., 1996). The link between histone deacetylation and repression was therefore established in the same way as the link between acetylation and activation – by showing that two proteins that were initially thought to have different activities are in fact related. These are good examples of the value of homology analysis in studies of gene and protein function (Section 7.2.1).

In common with HATs and nucleosome remodeling enzymes, HDACs are contained in multiprotein complexes. One of these is the mammalian Sin3 complex, which comprises at least seven proteins, including HDAC1 and HDAC2 along with others that do not have deacetylase activity but which provide ancillary functions essential to the process (Ng and Bird, 2000). Examples of ancillary proteins are RbAp46 and RbAp48, which are members of the Sin3 complex and are thought to contribute the histone-binding capability. RbAp46 and RbAp48 were first recognized through their association with the retinoblastoma protein, which controls cell proliferation by inhibiting expression of various genes until their activities are required and which, when mutated, leads to cancer (Brehm et al., 1998; Magnaghi-Jaulin et al., 1998). This link between Sin3 and a protein implicated in cancer provides a powerful argument for the importance of histone deacetylation in gene silencing. Other deacetylation complexes include NuRD in mammals, which combines HDAC1 and HDAC2 with a different set of ancillary proteins, and yeast Sir2 (Ahringer, 2000), which is different from other HDACs in that it has an energy requirement (Imai et al., 2000). The distinctive features of Sir2 show that HDACs are more diverse than originally realized, possibly indicating that novel roles for histone deacetylation are waiting to be discovered.

Studies of HDAC complexes are beginning to reveal links between the different mechanisms for genome activation and silencing. Both Sin3 and NuRD contain proteins that bind to methylated DNA (as described in the next section), and NuRD contains proteins that are very similar to components of the nucleosome remodeling complex Swi/Snf. NuRD does in fact act as a classical nucleosome remodeling machine in vitro. Further research will almost certainly unveil additional links between what we currently look on as different types of chromatin modification system, but which in reality may simply be different facets of a single grand design.

Box Icon

Box 8.2

Chromatin modification by the HMGN proteins. An additional dimension to chromatin modification is provided by the high mobility group N (HMGN) proteins. These are small proteins, each approximately 10 kDa, present in the nuclei of most vertebrates (more...)

Genome silencing by DNA methylation

Chromatin modification is not the only process that can bring about genome silencing. DNA methylation can also repress gene activity. Initially this was thought to be quite distinct from histone modification, but now we are beginning to see links between the two activities.

In eukaryotes, cytosine bases in chromosomal DNA molecules are sometimes changed to 5-methylcytosine by the addition of methyl groups by enzymes called DNA methyltransferases (see Figure 5.25). Cytosine methylation is relatively rare in lower eukaryotes but in vertebrates up to 10% of the total number of cytosines in a genome are methylated, and in plants the figure can be as high as 30%. The methylation pattern is not random, instead being limited to the cytosine in some copies of the sequences 5′–CG–3′ and, in plants, 5′–CNG–3′. Two types of methylation activity have been distinguished (Figure 8.11). The first is maintenance methylation which, following genome replication, is responsible for adding methyl groups to the newly synthesized strand of DNA at positions opposite methylated sites on the parent strand (Section 14.2.3). The maintenance activity therefore ensures that the two daughter DNA molecules retain the methylation pattern of the parent molecule. The second activity is de novo methylation, which adds methyl groups at totally new positions and so changes the pattern of methylation in a localized region of the genome. From the results of test-tube experiments, it was originally thought that the first DNA methyltransferase to be discovered, Dnmt1, was responsible for both types of methylation in mammalian cells. It was subsequently discovered that knockout mice (Section 7.2.2) that have an inactivated gene for Dnmt1 can still carry out de novo methylation. This led to the search for new enzymes and the eventual discovery of Dnmt3a and Dnmt3b (see Research Briefing 8.1), which are now considered to be the main de novo methylases of mammals, with Dnmt1 primarily responsible for the maintenance activity (Bird, 1999).

Box Icon

Box 8.1

Discovery of the mammalian DNA methyltransferases. Innovative experiments have identified the enzymes responsible for de novo methylation of mammalian DNA. Two DNA methylation activities are required in mammalian nuclei (see Figure 8.11). The first of (more...)

Methylation results in repression of gene activity (Jones, 1999). This has been shown by experiments in which methylated or unmethylated genes have been introduced into cells by cloning and their expression levels measured: expression does not occur if the DNA sequence is methylated. The link with gene expression is also apparent when the methylation patterns in chromosomal DNAs are examined, these showing that active genes are located in unmethylated regions. For example, in humans, 40–50% of all genes are located close to CpG islands (Section 7.1.1), with the methylation status of the CpG island reflecting the expression pattern of the adjacent gene. Housekeeping genes – those that are expressed in all tissues – have unmethylated CpG islands, whereas tissue-specific genes are unmethylated only in those tissues in which the adjacent gene is expressed. Note that because the methylation pattern is maintained after cell division, information specifying which genes should be expressed is inherited by the daughter cells, ensuring that in a differentiated tissue the appropriate pattern of gene expression is retained even though the cells in the tissue are being replaced and/or added to by new cells.

The importance of DNA methylation is underlined by studies of human diseases. The syndrome called ICF (immunodeficiency, centromere instability and facial anomalies) which, as the name suggests, has wide-ranging phenotypic effects, is associated with undermethylation of various genomic regions, and is caused by a mutation in the gene for Dnmt3b (Xu et al., 1999). The opposing situation – hypermethylation – is seen within the CpG islands of genes that exhibit altered expression patterns in certain types of cancer (Baylin and Herman, 2000), although in these cases the abnormal methylation could equally well be a result rather than the cause of the disease state.

How methylation influences genome expression was a puzzle for many years. Now it is known that methyl-CpG-binding proteins (MeCPs) are components of both the Sin3 and NuRD histone deacetylase complexes. This discovery has led to a model in which methylated CpG islands are the target sites for attachment of HDAC complexes that modify the surrounding chromatin in order to silence the adjacent genes (Figure 8.12).

Figure 8.12. A model for the link between DNA methylation and genome expression.

Figure 8.12

A model for the link between DNA methylation and genome expression. Methylation of the CpG island upstream of a gene provides recognition signals for the methyl-CpG-binding protein (MeCP) components of a histone deacetylase complex (HDAC). The HDAC modifies (more...)

Methylation is involved in imprinting and X inactivation

Further evidence, if it is needed, of the link between DNA methylation and genome silencing is provided by two intriguing phenomena called genomic imprinting and X inactivation.

Genomic imprinting is a relatively uncommon but important feature of mammalian genomes in which only one of a pair of genes, present on homologous chromosomes in a diploid nucleus, is expressed, the second being silenced by methylation (Feil and Khosia, 1999). It is always the same member of a pair of genes that is imprinted and hence inactive; for some genes this is the version inherited from the mother, and for other genes it is the paternal version. Thirty genes in humans and mice have been shown to display imprinting. An example is Igf2, which codes for a growth factor, a protein involved in signaling between cells (Section 12.1). In mice, only the paternal gene is active (Figure 8.13). On the chromosome inherited from the mother various segments of DNA in the region of Igf2 are methylated, preventing expression of this copy of the gene. Interestingly, a second imprinted gene, H19, is located some 90 kb away from Igf2, but the imprinting is the other way round: the maternal version of H19 is active and the paternal version is silent.

Figure 8.13. A pair of imprinted genes on human chromosome 11.

Figure 8.13

A pair of imprinted genes on human chromosome 11. Igf2 is imprinted on the chromosome inherited from the mother, and H19 is imprinted on the paternal chromosome. The drawing is not to scale: the two genes are approximately 90 kb apart.

The function of imprinting is not known. One possibility is that it has a role in development, because artificially created parthenogenetic mice, which have two copies of the maternal genome, fail to develop properly. More subtle explanations based on the evolutionary conflicts between the males and females of a species have also been proposed (Jaenisch, 1997).

X inactivation is much less enigmatic. This is a special form of imprinting that leads to total inactivation of one of the X chromosomes in a female mammalian cell (Heard et al., 1997). It occurs because females have two X chromosomes whereas males have only one. If both of the female X chromosomes were active then proteins coded by genes on the X chromosome might be synthesized at twice the rate in females compared with males. To avoid this undesirable state of affairs, one of the female X chromosomes is silenced and is seen in the nucleus as a condensed structure called the Barr body, which is comprised entirely of heterochromatin. Silencing occurs early in embryo development and is controlled by the X inactivation center, a discrete region present on each X chromosome. In a cell undergoing X inactivation, the inactivation center on one of the X chromosomes initiates the formation of heterochromatin, which spreads out from the nucleation point until the entire chromosome is affected, with the exception of a few short segments containing small clusters of genes that remain active. The process takes several days to complete. The exact mechanism is not understood but it involves, although is not entirely dependent upon, each of the following:

  • A gene called Xist, located in the inactivation center, which is transcribed into a 25-kb non-coding RNA, copies of which coat the chromosome as heterochromatin is formed;
  • Replacement of histone H2A, one of the members of the core octamer of the nucleosome (Section 2.2.1), with a special histone, macroH2A1 (Costanzi and Pehrson, 1998);
  • Deacetylation of histone H4 (Jeppesen and Turner, 1993), as usually occurs in heterochromatin;
  • Hypermethylation of certain DNA sequences, although this appears to occur after the inactive state has been set up.
    X inactivation is heritable and is displayed by all cells descended from the initial one within which the inactivation took place.

Study Aids For Chapter 8

Self study questions


How have developments in electron microscopy led to advances in our understanding of the internal structure of the eukaryotic nucleus?


Describe the differences between constitutive heterochromatin, facultative heterochromatin and euchromatin.


Distinguish between the terms ‘structural domain’ and ‘functional domain’. What experimental techniques are used to delineate a functional domain?


What are the special properties of the insulator sequences that mark the boundaries of functional domains?


Describe the important features of the β-globin locus control region.


Give examples of the various histone modifications that activate or silence regions of the genome. Explain what is meant by ‘histone code’.


Describe the features and activities of histone acetyltransferases and histone deacetylases.


Outline our current knowledge of the interactions between nucleosomes and HMGN proteins.


Using the Swi/Snf complex as an example, describe how nucleosome remodeling can influence gene expression


Write an essay on DNA methylation. Your essay should distinguish between the roles of maintenance and de novo methylation and include an account of the discovery of the mammalian methyltransferases.


What is meant by ‘genomic imprinting’?


Outline the key features of X inactivation.

Problem-based learning


To what extent can it be assumed that the picture of nuclear architecture built up by modern electron microscopy is an accurate depiction of the actual structure of the nucleus, as opposed to an artefact of the methods used to prepare cells for examination?


In many areas of biology it is difficult to distinguish between cause and effect. Evaluate this issue with regard to chromatin modifications and genome expression – do chromatin modifications cause changes in genome expression or are they the effect of these expression changes?


Explore and assess the histone code hypothesis.


Maintenance methylation ensures that the pattern of DNA methylation displayed by two daughter DNA molecules is the same as the pattern on the parent molecule. In other words, the methylation pattern, and the information on gene expression that it conveys, is inherited. Other aspects of chromatin structure might also be inherited in a similar way. How do these phenomena affect the Mendelian view that inheritance is specified by genes? (The 10 August 2001 issue of Science is a good starting point for your research into this topic.)


  1. Ahringer J. NuRD and SIN3: histone deacetylase complexes in development. Trends Genet. (2000);16:351–356. [PubMed: 10904264]
  2. Bannister AJ, Kouzarides T. The CBP co-activator is a histone acetyltransferase. Nature. (1996);384:641–643. [PubMed: 8967953]
  3. Bannister AJ, Zegerman P, Partridge JF. et al. Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature. (2001);410:120–124. [PubMed: 11242054]
  4. Baylin SB, Herman JG. DNA hypermethylation in tumorigenesis: epigenetics joins genetics. Trends Genet. (2000);16:168–174. [PubMed: 10729832]
  5. Bell AC, West AG, Felsenfeld G. Insulators and boundaries: versatile regulatory elements in the eukaryotic genome. Science. (2001);291:447–450. [PubMed: 11228144]
  6. Berger SL. Local or global? Nature. (2000);408:412–415. [PubMed: 11100706]
  7. Bird A. DNA methylation de novo. Science. (1999);286:2287–2288. [PubMed: 10636791]
  8. Bradbury EM. Reversible histone modifications and the chromosome cell cycle. BioEssays. (1992);14:9–16. [PubMed: 1312335]
  9. Brehm A, Miiska EA, McCance DJ, Reid JL, Bannister AJ, Kouzarides T. Retinoblastoma protein recruits histone deacetylase to repress transcription. Nature. (1998);391:597–601. [PubMed: 9468139]
  10. Brownell JE, Zhou J, Ranalli T. et al. Tetrahymena histone acetyltransferase A: a homolog to yeast Gcn5p linking histone acetylation to gene activation. Cell. (1996);84:843–851. [PubMed: 8601308]
  11. Bustin M. Chromatin unfolding and activation by HMGN chromosomal proteins. Trends Biochem. Sci. (2001);26:431–437. [PubMed: 11440855]
  12. Costanzi C, Pehrson JR. Histone macroH2A1 is concentrated in the inactive X chromosome of female mammals. Nature. (1998);393:599–601. [PubMed: 9634239]
  13. Feil R, Khosia S. Genomic imprinting in mammals: an interplay between chromatin and DNA methylation? Trends Genet. (1999);15:431–434. [PubMed: 10529801]
  14. Gerasimova TI, Byrd K, Corces VG. A chromatin insulator determines the nuclear localization of DNA. Mol. Cell. (2000);6:1025–1035. [PubMed: 11106742]
  15. Heard E, Clerc P, Avner P. X-chromosome inactivation in mammals. Ann. Rev. Genet. (1997);31:571–610. [PubMed: 9442908]
  16. Imai S, Armstrong CM, Kaeberlein M, Guarente L. Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature. (2000);403:795–800. [PubMed: 10693811]
  17. Jaenisch R. DNA methylation and imprinting: why bother? Trends Genet. (1997);13:323–329. [PubMed: 9260519]
  18. Jenuwein T, Allis CD. Translating the histone code. Science. (2001);293:1074–1080. [PubMed: 11498575]
  19. Jeppesen P, Turner BM. The inactive X chromosome in female mammals is distinguished by a lack of histone H4 acetylation, a cytogenetic marker for gene expression. Cell. (1993);74:281–289. [PubMed: 8343956]
  20. Jones PA. The DNA methylation paradox. Trends Genet. (1999);15:34–37. [PubMed: 10087932]
  21. Kellum R, Schedl P. A position-effect assay for boundaries of higher-order chromosomal domains. Cell. (1991);64:941–950. [PubMed: 1848159]
  22. Lachner M, O’Carroll D, Rea S, Mechtler K, Jenuwein T. Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature. (2001);410:116–120. [PubMed: 11242053]
  23. Lever MA, Th'ng JPH, Sun X, Hendzel MJ. Rapid exchange of histone H1. 1 on chromatin in living cells. Nature. (2000);408:873–876. [PubMed: 11130728]
  24. Lewis JD, Tollervey D. Like attracts like: getting RNA processing together in the nucleus. Science. (2000);288:1385–1389. [PubMed: 10827942]
  25. Li Q, Harju S, Peterson KR. Locus control regions: coming of age at a decade plus. Trends Genet. (1999);15:403–408. [PubMed: 10498936]
  26. Litt MD, Simpson M, Gaszner M, Allis CD, Felsenfeld G. Correlation between histone lysine methylation and developmental changes at the chicken β-globin locus. Science. (2001);293:2453–2455. [PubMed: 11498546]
  27. Lodish H, Berk A, Zipursky AL, Matsudaira P, Baltimore D and Darnell J (2000) Molecular Cell Biology, 4th edition. W. H. Freeman, New York.
  28. Magnaghi-Jaulin L, Groisman R, Naguibneva I. et al. Retinoblastoma protein represses transcription by recruiting a histone deacetylase. Nature. (1998);391:601–605. [PubMed: 9468140]
  29. Misteli T. Protein dynamics: implications for nuclear architecture and gene expression. Science. (2001);291:843–847. [PubMed: 11225636]
  30. Misteli T, Gunjan A, Hock R, Bustin M, Brown DT. Dynamic binding of histone H1 to chromatin in living cells. Nature. (2000);408:877–881. [PubMed: 11130729]
  31. Ng HH, Bird A. Histone deacetylases: silencers for hire. Trends Biochem. Sci. (2000);25:121–126. [PubMed: 10694882]
  32. Pennisi E. Opening the way to gene activity. Science. (1997);275:155–157. [PubMed: 8999545]
  33. Robzyk K, Recht J, Osley MA. Rad6-dependent ubiquitination of histone H2b in yeast. Science. (2000);287:501–504. [PubMed: 10642555]
  34. Strahl BD, Allis D. The language of covalent histone modifications. Nature. (2000);403:41–45. [PubMed: 10638745]
  35. Sudarsanam P, Winston F. The Swi/Snf family: nucleosome-remodelling complexes and transcriptional control. Trends Genet. (2000);16:345–351. [PubMed: 10904263]
  36. Syntichaki P, Topalidou I, Thireos G. The Gcn5 bromodomain co-ordinates nucleosome remodelling. Nature. (2000);404:414–417. [PubMed: 10746732]
  37. Taunton J, Hassig CA, Schreiber SL. A mammalian histone deacetylase related to the yeast transcriptional regulator Rpd3p. Science. (1996);272:408–411. [PubMed: 8602529]
  38. Tsien RY. The green fluorescent protein. Ann. Rev. Biochem. (1998);67:509–544. [PubMed: 9759496]
  39. Wolffe AP (1995) Genetic effects of DNA packaging. Sci. Am., Nov/Dec, 68–77.
  40. Xu G-L, Bestor TH, Bourc′his D. et al. Chromosome instability and immunodeficiency syndrome caused by mutations in a DNA methyltransferase gene. Nature. (1999);402:187–191. [PubMed: 10647011]

Further Reading

  1. Aalfs JD, Kingston RE. What does ‘chromatin remodelling’ mean? Trends Biochem. Sci. (2000);25:548–555.Stimulating discussion of histone modification and nucleosome remodeling. [PubMed: 11084367]
  2. Ballabio A, Willard HF. Mammalian X-chromosome inactivation and the XIST gene. Curr. Opin. Genet. Dev. (1992);2:439–447. [PubMed: 1504619]
  3. Brown CE, Lechner T, Howe L, Workman JL. The many HATs of transcription coactivators. Trends Biochem. Sci. (2000);25:15–19.Details of histone acetyltransferases and their mode of action. [PubMed: 10637607]
  4. Jones P A, Takai D. The role of DNA methylation in mammalian epigenetics. Science. (2001);293:1068–1070.A very readable account of our current knowledge of DNA methylation. [PubMed: 11498573]
  5. Lee TI, Young RA. Transcription of eukaryotic protein-coding genes. Ann. Rev. Genet. (2000);34:77–137.Includes extensive details of histone modification and nucleosome remodeling. [PubMed: 11092823]
  6. Roth SY, Denu JM, Allis CD. Histone acetyltransferases. Ann. Rev. Biochem. (2001);70:81–120. [PubMed: 11395403]
  7. Wu J, Grunstein M. 25 years after the nucleosome model: chromatin modifications. Trends Biochem. Sci. (2000);25:619–623.A useful review of this topic. [PubMed: 11116189]
Image ch1f14
Image ch2f7
Image permission
Image ch2f6
Image ch2f8
Image ch2f14
Image ch5f25
Copyright © 2002, Garland Science.
Bookshelf ID: NBK21137


Related information

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...