The Role of Methylation in Gene Expression

There are many ways that gene expression is controlled in eukaryotes, but methylation of DNA (not to be confused with histone methylation) is a common epigenetic signaling tool that cells use to lock genes in the "off" position. In recent decades, researchers have learned a great deal about DNA methylation, including how it occurs and where it occurs, and they have also discovered that methylation is an important component in numerous cellular processes, including embryonic development, genomic imprinting, X-chromosome inactivation, and preservation of chromosome stability. Given the many processes in which methylation plays a part, it is perhaps not surprising that researchers have also linked errors in methylation to a variety of devastating consequences, including several human diseases.

Prior to 1980, there were a number of clues that suggested that methylation might play a role in the regulation of gene expression. For example, J. D. McGhee and G. D. Ginder compared the methylation status of the beta-globin locus in cells that did and did not express this gene. Using restriction enzymes that distinguished between methylated and unmethylated DNA, the duo showed that the beta-globin locus was essentially unmethylated in cells that expressed beta-globin but methylated in other cell types (McGhee & Ginder, 1979). This and other evidence of the time were indirect suggestions that methylation was somehow involved in gene expression.

Shortly after McGhee and Ginder published their results, a more direct experiment that examined the effects of inhibiting methylation on gene expression was performed using 5-azacytidine in mouse cells. 5-azacytidine is one of many chemical analogs for the nucleoside cytidine. When these analogs are integrated into growing DNA strands, some, including 5-azacytidine, severely inhibit the action of the DNA methyltransferase enzymes that normally methylate DNA. (Interestingly, other analogs, like Ara-C, do not negatively impact methylation.) Because most DNA methylation was known to occur on cytosine residues, scientists hypothesized that if they inhibited methylation by flooding cellular DNA with 5-azacytidine, then they could compare cells before and after treatment to see what impact the loss of methylation had on gene expression. Knowing that gene expression changes are responsible for cellular differentiation, these researchers used changes in cellular phenotypes as a proxy for gene expression changes (Table 1; Jones & Taylor, 1980).

Table 1: Effect of Cytidine Analogs on Cell Differentiation and DNA Methylation

Today, researchers know that DNA methylation occurs at the cytosine bases of eukaryotic DNA, which are converted to 5-methylcytosine by DNA methyltransferase (DNMT) enzymes. The altered cytosine residues are usually immediately adjacent to a guanine nucleotide, resulting in two methylated cytosine residues sitting diagonally to each other on opposing DNA strands. Different members of the DNMT family of enzymes act either as de novo DNMTs, putting the initial pattern of methyl groups in place on a DNA sequence, or as maintenance DNMTs, copying the methylation from an existing DNA strand to its new partner after replication. Methylation can be observed by staining cells with an immunofluorescently labeled antibody for 5-methylcytosine. In mammals, methylation is found sparsely but globally, distributed in definite CpG sequences throughout the entire genome, with the exception of CpG islands, or certain stretches (approximately 1 kilobase in length) where high CpG contents are found. The methylation of these sequences can lead to inappropriate gene silencing, such as the silencing of tumor suppressor genes in cancer cells.

Currently, the mechanism by which de novo DNMT enzymes are directed to the sites that they are meant to silence is not well understood. However, researchers have determined that some of these DNMTs are part of chromatin-remodeling complexes and serve to complete the remodeling process by performing on-the-spot DNA methylation to lock the closed shape of the chromatin in place.

The roles and targets of DNA methylation vary among the kingdoms of organisms. As previously noted, among Animalia, mammals tend to have fairly globally distributed CpG methylation patterns. On the other hand, invertebrate animals generally have a "mosaic" pattern of methylation, where regions of heavily methylated DNA are interspersed with nonmethylated regions. The global pattern of methylation in mammals makes it difficult to determine whether methylation is targeted to certain gene sequences or is a default state, but the CpG islands tend to be near transcription start sites, indicating that there is a recognition system in place.

Plantae are the most highly methylated eukaryotes, with up to 50% of their cytosine residues exhibiting methylation. Interestingly, in Fungi, only repetitive DNA sequences are methylated, and in some species, methylation is absent altogether, or it occurs on the DNA of transposable elements in the genome. The mechanism by which the transposons are recognized and methylated appears to involve small interfering RNA (siRNA). The whole silencing mechanism invoking DNMTs could be a way for these organisms to defend themselves against viral infections, which could generate transposon-like sequences. Such sequences can do less harm to the organism if they are prevented from being expressed, although replicating them can still be a burden (Suzuki & Bird, 2008). In other fungi, such as fission yeast, siRNA is involved in gene silencing, but the targets include structural sequences of the chromosomes, such as the centromeric DNA and the telomeric repeats at the chromosome ends.

For many years, methylation was believed to play a crucial role in repressing gene expression, perhaps by blocking the promoters at which activating transcription factors should bind. Presently, the exact role of methylation in gene expression is unknown, but it appears that proper DNA methylation is essential for cell differentiation and embryonic development. Moreover, in some cases, methylation has observed to play a role in mediating gene expression. Evidence of this has been found in studies that show that methylation near gene promoters varies considerably depending on cell type, with more methylation of promoters correlating with low or no transcription (Suzuki & Bird, 2008). Also, while overall methylation levels and completeness of methylation of particular promoters are similar in individual humans, there are significant differences in overall and specific methylation levels between different tissue types and between normal cells and cancer cells from the same tissue.

Researchers have also determined that mice that lack a particular DNMT have reduced methylation levels and die early in development (Suzuki & Bird, 2008). This is not the case for all eukaryotes, however; some organisms, such as the yeast Saccharomyces cerevisiae and the nematode worm Caenorhabditis elegans, are thought to have no methylated DNA at all (although, at least in yeast, there are sequences in their genomes that are homologous to those that code for the DNMT enzymes).

Although patterns of DNA methylation appear to be relatively stable in somatic cells, patterns of histone methylation can change rapidly during the course of the cell cycle. Despite this difference, several studies have indicated that DNA methylation and histone methylation at certain positions are connected. For instance, results of immunoprecipitation studies using human cells suggest that DNA methylation and histone methylation work together during replication to ensure that specific methylation patterns are passed on to progeny cells (Sarraf & Stancheva, 2004). Indeed, evidence has been presented that in some organisms, such as Neurospora crassa (Tamaru & Selker, 2001) and Arabidopsis thaliana (Jackson et al., 2002), H3-K9 methylation (methylation of a specific lysine residue in the histone H3) is required in order for DNA methylation to take place. However, exceptions have been observed in which the relationship is reversed. In one study, for example, H3 methylation was reduced at a tumor suppressor gene in cells deficient in DNA methyltransferase (Martin & Zhang, 2005).

In an interestingly coordinated process, proteins that bind to methylated DNA also form complexes with the proteins involved in deacetylation of histones. Therefore, when DNA is methylated, nearby histones are deacetylated, resulting in compounded inhibitory effects on transcription. Likewise, demethylated DNA does not attract deacetylating enzymes to the histones, allowing them to remain acetylated and more mobile, thus promoting transcription.

In most cases, methylation of DNA is a fairly long-term, stable conversion, but in some cases, such as in germ cells, when silencing of imprinted genes must be reversed, demethylation can take place to allow for "epigenetic reprogramming." The exact mechanisms for demethylation are not entirely understood; however, it appears that this process may be mediated by the removal of amino groups by DNA deaminases (Morgan et al., 2004). After deamination, the DNA has a mismatch and is repaired, causing it to become demethylated. In fact, studies using inhibitors of one DNMT enzyme showed that this enzyme was involved in not only DNA methylation, but also in the removal of amino groups.

Given the critical role of DNA methylation in gene expression and cell differentiation, it seems obvious that errors in methylation could give rise to a number of devastating consequences, including various diseases. Indeed, medical scientists are currently studying the connections between methylation abnormalities and diseases such as cancer, lupus, muscular dystrophy, and a range of birth defects that appear to be caused by defective imprinting mechanisms (Robertson, 2005). The results of these studies will be invaluable for treating these disorders, as well as for understanding and preventing complications that can arise during embryonic development due to abnormalities in X-chromosome methylation and gene imprinting.

To date, a large amount of research on DNA methylation and disease has focused on cancer and tumor suppressor genes. Tumor suppressor genes are often silenced in cancer cells due to hypermethylation. In contrast, the genomes of cancer cells have been shown to be hypomethylated overall when compared to normal cells, with the exception of hypermethylation events at genes involved in cell cycle regulation, tumor cell invasion, DNA repair, and others events in which silencing propagates metastasis (Figure 1; Robertson, 2005). In fact, in certain cancers, such as that of the colon, hypermethylation is detectable early and might serve as a biomarker for the disease.

Figure 1: DNA methylation and cancer.

The diagram shows a representative region of genomic DNA in a normal cell. The region shown contains repeat-rich, hypermethylated pericentromeric heterochromatin and an actively transcribed tumour suppressor gene (TSG) associated with a hypomethylated CpG island (indicated in red). In tumour cells, repeat-rich heterochromatin becomes hypomethylated and this contributes to genomic instability, a hallmark of tumour cells, through increased mitotic recombination events. De novo methylation of CpG islands also occurs in cancer cells, and can result in the transcriptional silencing of growth-regulatory genes. These changes in methylation are early events in tumorigenesis.

© 2005 Nature Publishing Group Robertson, K. DNA methylation and human disease. Nature Reviews Genetics 6, 598. All rights reserved.

Within the past thirty years, researchers have discovered numerous details about the process of DNA methylation. For instance, scientists now know that methylation plays a critical role in the regulation of gene expression, and they have also determined that this process tends to occur at certain locations within the genomes of different species. Furthermore, DNA methylation has been shown to play a vital role in numerous cellular processes, and abnormal patterns of methylation have been liked to several human diseases. Nonetheless, as with other topics in the field of epigenetics, gaps remain in our knowledge of DNA methylation. As new laboratory techniques are developed and additional genomes are mapped, scientists will no doubt continue to uncover many of the unknowns of how, when, and where DNA is methylated, and for what purposes.