The Role Of DNMT3B DNA Methyltransferase Gene Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Epigenetics, in biology, is the study of inherited variations in gene expression that can modulate the use of the genome during the development of complex organisms, and these are caused by mechanisms that do not produce changes at the underlying DNA sequence level. Epigenetic modifications primarily cause variations in the chromatin structure of the DNA and such alteration processes include DNA methylation, post-translational histone tail modifications and non-histone proteins that bind to the chromatin structure. These epigenetic modification processes are implicated in other subsequent procedures that include cell reprogramming and gene silencing, genomic imprinting and X-chromosome inactivation etc (Morgan et al, 2005; Bernstein et al, 2007). In eukaryotic biology, the best example of an epigenetic variation is cellular differentiation, where during morphogenesis the single fertilized zygote differentiates into many cell types, such as muscle and nerve cells, blood vessels etc. (Fig. 1), and it accomplishes this by activating particular genes while silencing others.

Figure 1: Cell cycle. During morphogenesis, the single fertilized zygote divides to form totipotent stem cells which will become the different pluripotent cell lines of the embryo and finally form fully differentiated cells. Picture adapted from "Latest Stem Cells News" on 06/04/10;

During the development of complex multicellular organisms, different cells and tissues acquire diverse programmes of lineage-specific gene expression through these regulated epigenetic processes. Therefore, every cell and tissue of an organism will have its own unique epigenome, which refers to the overall epigenetic state of its cells and tissues, and this will be reflected both in the genotype and the phenotype of the particular organism (Fig. 2) (Morgan et al, 2005; Bernstein et al, 2007). In principal, each cell or tissue of an organism contains the same genetic information, but there is selective usage of that genetic information.

Figure 2: Epigenome. An organism's epigenome within a particular cell is a result of the combinatory function of not only genetic determinants, but also of lineage-specific and environmental signals. Adapted from Bernstein et al., 2007.

Epigenetic modifications commonly take place early during cell differentiation and usually become fixed once the cell has completely differentiated or exited the cell cycle, forming the organisms' specific cell memory. Consequently, these epigenetic changes are preserved every time the cells divide (Bernstein et al, 2007). There are occasions, however, either in normal development or during some disease situations, when major epigenetic reprogramming occurs within the nucleus of an organism (Fig. 3). This is most frequently undertaken by cells when their developmental potency has been altered. During this reprogramming process, epigenetic marks which have been fixed into the epigenetic code of the organism during early development are removed and replaced with a different set of marks. Epigenetic alterations are nevertheless reversible. There are some other cases, i.e. if a mutation has changes the DNA sequence (known as epimutations; Siedlecki & Zielenkiewicz, 2006), when some epigenetic variations may be inherited from one generation to the next (Morgan et al, 2005).

Figure 3: Epigenetic reprogramming cycle. Epigenetic modifications undergo reprogramming during the life cycle in two phases: during gametogenesis and preimplantation development. Adapted from Morgan et al., 2005.

In multicellular organisms, gene expression can be regulated through different epigenetic processes, which take place so as to confer gene expression stability during the organism's development (Zhu, 2009). One of these processes, which plays a major role in the regulation of genes, is chromatin remodeling and this is achieved through two main mechanisms, i.e. DNA methylation and post-translational histone modifications, such as histone acetylation, phosphorylation and ubiquitination (Morgan et al, 2005; Bernstein et al, 2007). From these two mechanisms of chromatin remodeling, DNA methylation is the most wide spread epigenetic alteration that can occur to the DNA sequence and the way by which it is achieved during differentiation is the main scope of interest of this research study.

DNA methylation, in general, is the biochemical process by which a methyl group is covalently added to the DNA sequence either at the 5' carbon position of a cytosine pyrimidine ring (in all organisms; mammals), or otherwise at the 5' carbon of the cytosine of a CpG dinucleotide (Fig. 4, A), or at the 6' nitrogen position of an adenine purine ring (in almost all eukaryotes and in bacteria). This epigenetic process is vital for normal development and cellular differentiation in complex organisms, and it promotes the alteration of gene expression patterns in a stable way further accomplishing cellular memory (Siedlecki & Zielenkiewicz, 2006). DNA methylation is also associated with a number of other important epigenetic processes, including X-chromosome inactivation and imprinting, suppression of viral gene expression, as well as silencing other deleterious for the organism elements, developmental disorders, aging and carcinogenesis (Suetake et al, 2004; Bernstein et al, 2007; Zhu, 2009).



Figure 4: A. Methylation by DNA methyltransferases at CpG islands. B. DNA demethylation relaxes chromatin structure allowing histone acetylation and the binding of transcriptional complexes. Modified from Bernstein et al., 2007.

At fertilization, as the parental genomes are in different stages of the cell cycle and hence also have quite diverse epigenetic marks and chromatin organization, gametic DNA methylation is generally removed, so as to be replaced with embryonic marks essential for early development, toti- and pluri- potency, and only re-established through successive rounds of cellular division during development. More specifically, during zygote formation the paternal genome is found in a DNA demethylated state, having exchanged protamines for histones in the male pronucleus which have become readily modified, while the maternal genome appears more static epigenetically with the histones having been progressively modified only to a small extent (Bernstein et al, 2007). Paternal genome demethylation is active and it involves the presence of DNA demethylases, while maternal genome demethylation is passive (Fig. 3; Morgan et al, 2005).

During the early preimplantation development stage, passive DNA demethylation takes place together with additional reorganization of histone modifications, steadily leading to de novo DNA methylation after implantation has occurred (Fig. 3). Despite DNA demethylation during the preimplantation development stage, imprinted genes are able to maintain their methylation status through the preimplantation reprogramming process. Imprinting marks originating from the parental genomes, which are specifically located within the sperm and the oocyte, are protected from this genome-wide reprogramming process (Morgan et al, 2005).

During blastulation, the inner cell mass (ICM) is hypermethylated compared to the trophectoderm (TE) (Bernstein et al, 2007). This epigenetic difference in the DNA methylation state between the first two lineages of the blastocyst stage, may represent the importance of epigenetic reprogramming required for cell differentiation, accurate embryonic gene expression and early lineage development. In the early embryo, the genome of the primordial germ cells (PGCs) undergoes DNA demethylation. Following this loss of methylation, the genomes of the gametes are de novo DNA methylated and acquire sex-specific imprints (Fig. 3; Morgan et al, 2005).

In adult somatic cells and tissues, DNA methylation characteristically occurs in the context of CpG dinucleotides. These sequences only represent approximately 1% of the human genome. In mammals, between 70% to 80% of the CpGs are methylated, and these are found in clusters named 'CpG islands' the majority of which remain unmethylated during normal development, and which are typically present within regulatory sequences; at the 5' untranslated regulatory region of different genes, in close proximity to their promoters, or around gene transcription start sites (Bird, 2002; Siedlecki & Zielenkiewicz, 2006; Bernstein et al, 2007). DNA methylation has an important function in the regulation of gene transcription, controlling and impeding this process in two different ways. On the one hand, methylation of the DNA sequence may itself render impossible the binding of essential transcriptional proteins to the specific gene, therefore blocking transcription, while, on the other hand, methylated DNA may recruit and bind to proteins known as methyl-CpG-binding domain proteins (MBDs). These proteins function by recruiting other proteins to the particular area on the gene sequence, including histone deacetylases and additional chromatin remodeling proteins, thus forming a compact, silent chromatin structure that is inactive. As a result, transcription of the gene is once again blocked (Fig. 4, B) (Bird, 2002; Burgers et al, 2002; Bernstein et al, 2007).

In mammals, the process of DNA methylation is undertaken with the aid of the DNA methyltransferase genes. These genes are distantly related to one another, probably having diverged during early evolution, and have different types of enzymatic activity; they can either maintain the methylation state of the genome or cause de novo methylation. There are specifically three main DNA methyltransferase (Dnmts) enzymes: Dnmt1, Dnmt3a and Dnmt3b (Burgers et al, 2002; Suetake et al, 2004). An additional Dnmt enzyme, whose function has not been entirely established yet, is the DNA methyltransferase 2 (Dnmt2) and its isoform Dnmt2a. These proteins are the most prevalent out of all the Dnmt enzymes, being expressed in most mouse and human adult tissues, and even though they appear to be DNA methyltransferase homologs, they do not methylate DNA but rather they seem to methylate small tRNAs, thus making them RNA methyltransferases. They have also been shown to lack the N-terminal regulatory domain, with only the C-terminal catalytic domain being conserved within their protein structure (Siedlecki & Zielenkiewicz, 2006).

These main Dnmt enzymes have been demonstrated to work alongside with histone deacetylases (HDACs), each in a different way, causing repression of gene transcription and hence gene silencing. However, eukaryotic Dnmts do not exhibit any sequence specificity apart from their ability to bind to CpG dinucleotides. Consequently, it is thought that the Dnmt enzymes are targeted to specific promoters or other genomic regions through protein-protein interactions, i.e. with particular DNA-binding proteins, such as methylated-histone binding proteins, transcription factors and co-repressors, leading to gene silencing. In other words, binding of Dnmts to particular genomic regions most possibly entails chromatin modifications and the presence of other remodeling proteins, which render the DNA sequence accessible to the methyltransferase enzymes (Bird, 2002; Burgers et al, 2002; Bernstein et al, 2007).

DNA methyltransferase 1 (Dnmt1) has been proposed to be involved in the maintainance of genomic DNA methylation by preserving the methylation patterns after every cellular DNA replication cycle, thus daughter cells inherit their DNA methylation patterns from the parental cells (Bird, 2002; Bernstein et al, 2007; Burgers et al, 2002). In addition, the Dnmt1 protein has been discovered to possess the ability to methylate substrates de novo (Burgers et al, 2002). This Dnmt enzyme has also been determined to be essential for genomic imprinting and inactivation of the X-chromosome in the mammalian embryo (Siedlecki & Zielenkiewicz, 2006). This methyltransferase family has been found to include other isoforms; apart from the somatic Dnmt1 methyltransferase (Dnmt1s) enzyme, a splice variant (Dnmt1b) and an oocyte-specific isoform (Dnmt1o) have also been determined (Morgan et al, 2005). These isoforms most probably have similar functions, although the exact way by which they all exert their effect is not quite well understood.

Studies have shown that the maintenance Dnmt1 protein works through a semi-conservative copying process by which parental-strand methylation patterns are inherited between cell generations. Dnmt1 specifically functions by hemi-methylating newly emerged CpG dinucleotides whose partners on the complementary parental strand already carry a methyl group (Bird, 2002; Burgers et al, 2002; Siedlecki & Zielenkiewicz, 2006; Bernstein et al, 2007). The Dnmt1o enzyme also appears to have an important role in the maintenance of DNA methylation during cell development. Particularly, during the epigenetic reprogramming cycle (Fig. 3), while passing from the preimplantational development stage to the blastocyst stage, the oocyte inherited Dnmt1 protein is initially involved in causing passive DNA demethylation. This takes place at the very early phases of cell division and is a result of the Dnmt1o protein being excluded from the nucleus of the cell. After further cell divisions, Dnmt1o enters back into the nucleus of the cell where it helps to maintain imprinted DNA methylation (Morgan et al, 2005).

DNA demethylation, which passively occurs in the presence of the Dnmt1 enzyme during preimplantation reprogramming, is a mechanism vital for a number of cellular processes during development, disease and defense of organisms (Morgan et al, 2005). This procedure has been shown to be mediated either by direct removal of methyl groups from a DNA sequence or through a base excision repair (BER) machinery. Moreover, it is accomplished in the presence of DNA demethylase enzymes, which are required for the removal of any undesirable DNA methylation that has been generated by promiscuous de novo methyltransferases or DNA methylation remodeling/reprogramming produced in response to environmental cues during development (Fig. 5; Zhu, 2009).

Figure 5: Establishment of DNA methylation patterns through the co-operation of DNA methyltransferases and demethylases. Adapted from Zhu, 2009.

The DNA methylatransferase 3 (Dnmt3) family of enzymes are structurally similar to the Dnmt1 class, with their C-terminal catalytic domain being attached to their N-terminal regulatory region through few Gly-Lys dipeptide repeats. The Dnmt3 family particularly consists of three isoforms: Dnmt3a, Dnmt3b and Dnmt3L (Siedlecki & Zielenkiewicz, 2006). It has been suggested that the Dnmt3a and Dnmt3b enzymes function as the de novo methyltransferases, meaning that these proteins can attach methyl groups to unmethylated DNA, and these recognize signals in the DNA that allow them to newly methylate cytosines, therefore arranging the DNA methylation patterns early during development (Burgers et al, 2002). These two Dnmt3s may interact and co-operate with the Dnmt1 enzyme during methylation events. However, Dnmt3a methylates CpG sites at a much slower rate than Dnmt1, but at a much greater rate than Dnmt3b, while Dnmt3a has also been found to methylate non-CpG sites. By working together, these two families, Dnmt1 and Dnmt3, can generate new DNA methylation patterns which are maintained in somatic cells through cell divisions (Bird, 2002; Siedlecki & Zielenkiewicz, 2006).

From the two Dnmt3 isoforms, more is known about Dnmt3b, which has been shown to play a very important role in early de novo methylation. Dnmt3b is especially required for the methylation of particular genomic regions, such as CpG islands on inactive X-chromosomes and pericentromeric repetitive sequences (Bird, 2002; Siedlecki & Zielenkiewicz, 2006). During the blastocyst stage of cell development (Fig. 3), Dnmt3b has been detected in a great degree in the ICM, leading to the extensive de novo DNA methylation that is observable. Its absence from the TE may explain the low levels of methylation in that area of the blastocyst (Morgan et al, 2005). Other studies have argued that the Dnmt3b protein may be capable of methylating regions of silent chromatin (Bird, 2002). However, the presence of both Dnmt3 enzymes is necessary for proper mammalian embryo development.

DNA methyltransferase 3-like (Dnmt3L), expressed mostly in germ cells, is a protein homolog of the Dnmt3 class of methyltransferase enzymes, but it is not a functional methyltransferase as it lacks several conserved motifs within its catalytic domain, i.e. the ATRX and PWWP domain (Siedlecki & Zielenkiewicz, 2006). Nevertheless, it is still capable of interacting with HDACs hence suppressing gene transcription (Burgers et al, 2002). Dnmt3L has particularly been found to be essential for the organization of the maternal genomic imprints early in development, even though it lacks any catalytic activity, and also for retaining the methylation of satellite DNA sequences at normal levels (Ehrlich et al, 2008). Its exact role is to interact with the other Dnmt3 isoforms by co-localizing in the nucleus, and in this way it assists their binding to specific DNA sequences by increasing their affinity to bind and stimulating their activity. The Dnmt3L protein together with the other two Dnmt3s specifically interact through their C-terminal regions, nevertheless its interaction with Dnmt3a is much stronger than what it is with Dnmt3b (Suetake et al, 2004).

A number of developmental aberrations and diseases have been verified to be related either with defective or absent DNA methyltransferases, especially with the Dnmt3 genes, confirming the importance of these enzymes for organism development (Siedlecki & Zielenkiewicz, 2006). An example of such a neurodevelopmental disorder is ICF (or namely immunodeficiency, centromeric instability and facial dysmorphism) syndrome, which has been shown to involve biallelic mutation, mostly single-base and missence, in the catalytic domain of the de novo Dnmt3b gene causing loss of gene activity and hence hypomethylation of the satellite 2 and 3 DNA (Sat2 & Sat3) in a small part of the patients genome (Ehrlich, 2002; Ueda et al, 2006). Consequently, as most of the ICF mutations arise in the coding C-terminal domain of the Dnmt3b gene, ICF is most possibly caused due to the lack of Dnmt3b activity rather than due to alteration in Dnmt3b protein-protein interactions, which usually occur outside the catalytic C-terminal domain. The loss of gene activity, in this case, is not complete but most likely partial, and thus, the ICF-causing mutations in Dnmt3b must leave some residual activity, otherwise embryonic lethality would most probably occur (as seen from in vivo mouse models). However, mutations within this methyltransferase gene are not such a common aspect of the ICF syndrome, as they are not always demonstrated in the patients investigated, suggesting either that mutations may exist outside of the coding Dnmt3b gene sequence or that other genes are quite likely to be engaged in the appearance of the disease, thus further suggesting genetic heterogeneity (Wijmenga et al, 2000; Hagleitner et al, 2007; Ehrlich et al, 2008).

The human Dnmt3b locus found to be associated with the disease has been mapped to the proximal long arm of chromosome 20 at position q11.21 (Fig. 6) (Wijmenga et al, 2000; Hagleitner et al, 2007). Additionally, the Dnmt3b gene has been shown to consist of a total of 23 exons spanning a genomic region of approximately 50 kb (Wijmenga et al, 2000), while to date eight different transcripts of the human Dnmt3b gene have been reported.

Figure 6: Genomic location of Dnmt3b gene, i.e. 20q11.2, (Chr 20: 31,350,191-31,397,162). Modified from Gene Cards®: The Human Gene Compendium on 14/04/10;

In mice, experiments have revealed that the Dnmt3b gene is also vital for normal mouse embryo development, as well as for de novo methylation of mouse embryonic stem (ES) cells. In particular, insertional inactivation of Dnmt3b, which specifically causes the disruption of both Dnmt3b alleles, results in prenatal death very early after implantation. This finding implies that complete loss of this gene is incompatible with life, and therefore, would lead to spontaneous abortions (Wijmenga et al, 2000; Ehrlich et al, 2008). The embryonic lethality of Dnmt3b null mice may either be due to the repression of genes that control cell growth or to the dysregulated activation of such genes that could cause improper growth arrest (Ueda et al, 2006). The study conducted by Ueda et al. (2006) revealed that Dnmt3b is moreover essential for the development of the liver, fetal heart and craniofacial features, and for the proper survival of T-cells, which are dysfunctional and apoptotic in ICF patients, thus helping to maintain them at normal levels. Dnmt3b was also discovered to be highly expressed in the placental tissues. As a result, the various cardiovascular defects, as well as the placental defects, may also be the cause of embryonic lethality. The ICF-like mutant mouse model demonstrated a somewhat similar phenotype to that of the human ICF syndrome. As a consequence, the creation of an ICF-like syndrome mouse model is of great significance due to the limitation in the number of ICF patients found worldwide, thus facilitating in the study of ICF symptoms (Ueda et al, 2006). Hopefully, such a model will help and assist our greater understanding of the ICF syndrome.

ICF, which was first described in the late 1970's, is a very rare autosomal recessive immune syndrome characterized by immunoglobin deficiency, branching of chromosomes 1, 9 and 16 and facial anomalies, and to date has only been described in roughly 50 patients worldwide (Ehrlich et al, 2008). Immunodeficiency is attributed to extremely reduced serum immunoglobin levels (i.e. hypo- and agamma- globulinaemia), leading to recurrent and prolonged infections. The nature of the immunodeficiency, nonetheless, is variable with different patients demonstrating different types of immunoglobulin protein deficiencies; either a single immunoglobulin deficiency or a combined immunodeficiency (Wijmenga et al, 2000). This immune deficiency is most likely caused due to the Dnmt3b interfering with lymphogenesis or lymphocyte activation (Hagleitner et al, 2007; Ueda et al, 2006).

Centromeric instability is the hallmark of the disease and is specific to the pericentromeric regions of chromosomes 1, 9 and 16 (Fig. 7), where DNA hypomethylation is strongly diminished, leading to decondansation of those chromosomal regions which become prone to breakage. Rejoining of these broken chromosomes leads to the formation of multiradiate chromosomes. These products are observed upon cytogenetic testing (i.e. after phytohaemagglutinin (PHA)-mitogen stimulation of lymphocytes) of ICF patients (Wijmenga et al, 2000; Ehrlich, 2002; Ueda et al, 2006).

Figure 7: Hypomethylated DNA in the juxtacentromeric heterochromatin (qh) regions of the ICF-implicated chromosomes. Modified from Ehrlich et al., 2008.

Facial anomalies are a common feature of ICF patients and are usually mild. These include characteristics such as epicanthic folds, hypertelorism, low set ears, broad flat nasal bridge, telecanthus, micrognathia and macroglossia with a protruding tongue. Congenital defects, such as cleft palate, syndactyly, hypospadia, hypothyroidism, have also been reported in some patients, as well as some cardiac anomalies and cerebral malformations. Haematological malignancies have also been uncovered in a few patient cases. Mental and phychomotor retardation are quite uncommon features of ICF patients, nevertheless can occasionally be observed (Hagleitner et al, 2007; Ehrlich et al, 2008).

ICF patients have a poor life expectancy due to the severe opportunistic infections (also revealing a reduced T-cell count), gastrointestinal problems, including severe diarrhea, and failure to thrive, all of which become apparent early during childhood. However, the immunodeficiency in ICF patients is the major contributor to the elevated mortality rate detected in early childhood, and could possibly be explained due to the alteration of transcription factor binding to the disease-related hypomethylated satellite DNA (Ehrlich et al, 2008). Immune infections typically include respiratory tract and pulmonary infections. Early prenatal diagnosis of the disease is of great importance as immunoglobin supplementation can significantly improve the patient's life (Hagleitner et al, 2007).

The purpose of this project is to comprehend the exact way by which the epigenome is formed during embryonic cell differentiation. For this purpose, DNA methylation patterns, and the way these are achieved during somatic differentiation, are the primary subject of concern and investigation. To answer our questions, we will initially study the DNA methyltransferase protein-protein interactions and how these methylase genes are targeted to their particular genomic sites. For this reason, the specific mouse Dnmt3b mutations (i.e. mouse equivalent mutations corresponding to the characterized human ones found in ICF syndrome patients) will be created and introduced into the coding sequence of the Dnmt3b gene, so as to further test the functionality of the Dnmt3b protein in mouse ES cells. In this way, we are hoping to dissect the Dnmt3b specific and redundant functions and gene targets in mouse ES cells during differentiation.