DNA methylation is an important epigenetic process where a methyl group is added to the cytosine or guanine nucleotides in the genome. DNA methyltransferases represent the key enzymes in such modification of the DNA, which is involved in gene transcription control, maintenance of genome stability and parental imprinting.
Those enzymes transfer a methyl group from S-adenosylmethionine to the nucleotide residues, and can be used to generate methylated DNA at specific sites. Global cytosine methylation patterns in mammals appear to be controlled by a complex interaction of independently encoded DNA methyltransferases: DNMT1, DNMT2, DNMT3A and DNMT3B.
DNMT1 is the most abundant methyltransferase in adult somatic cells. It attaches to hemi-methylated DNA (i.e. DNA with only one stand methylated) at CpG sites. It is often referred to as maintenance methyltransferase since it is thought to be the primary enzyme in charge of copying methylation patterns after DNA replication.
Following DNA replication the parent strand remains methylated, while the newly synthesized strand is not. DNMT1 then binds to hemi-methylated CpG sites on the newly synthesized strain and there it methylates the cytosine, thus maintaining established CpG methylation pattern through mitosis.
The experiments on DNMT1-knockout mice have revealed that DNMT1 is essential for proper early embryo development, imprinting, and X-chromosome inactivation. It has also been established that overexpression of DNMT1 serves as a hallmark in malignancies, such as endometrioid and prostate cancers.
Although DNMT2 shares a strong sequence homology to the other methyltransferases, it shows barely detectable DNA-cysteine methylation activity; hence its exact role in the DNA methylation process remains unclear. It has low expression levels in all the tissues, and inactivation of the homologous DNMT2 gene in embryonic stem cells of a mouse did not change the maintenance and methylation of DNA.
Still, it was demonstrated that DNMT2 could readily catalyze RNA methylation. Given that both DNA and RNA can serve as substrates for this enzyme, and that the affinity of DNA to DNMT2 is lower than RNA, it is believed that DNMT2 could be an evolutionary product where methyltransferases adjusted from a DNA to an RNA target.
In addition, methylation of tRNAs has an impact on the stability and folding of their structure. Such methylation of tRNAs may serve a protective function in mammals. Genes coding for similar DNA methylase were also detected in fungi and even in plants.
DNMT3a and DNMT3b are also known as “de novo” methyltransferases, since they do not require hemi-methylated DNA to bind, and they show an equivalent affinity for hemi-methylated and non-methylated DNA. Both enzymes are essential for the early development, and the loss of either of them is lethal due to a massive instability of chromosomes.
The DNMT3 family also includes a catalytically inert member produced during gametogenesis, known as DNMT3L. This enzyme is crucial for proper development, and when it is bound to either DNMT3a or DNMT3b, it can increase their catalytic activity 15-fold.
Overexpression of enzymes from this family is associated with carcinogenesis, reflecting that both DNMT3a and DNMT3b have their own unique set of functions or characteristics. Mutation of DNMT3b (and not DNMT3a) has a strong correlation with immunodeficiency, facial anomalies syndrome and centromeric instability.
It is probable that all three DNMTs possess both maintenance and “de novo” functions in vivo, hence they cannot be functionally divided. DNMT1 can sometimes function as a “de novo” DNMT and its overexpression can lead to “de novo” methylation of CpG islands. Likewise, DNMT3a and DNMT3b can fill the role of a maintenance enzyme.
Cooperation among DNMT1, DNMT3a and DNMT3b in maintenance of DNA methylation is also proven. Such collaboration among DNMTs is found in carcinogenesis as well, although their expression is diverse in various types of tumors.
It was recently demonstrated that DNMTs can also act as demethylases; DNMT1, DNMT3a and DNMT3b are all able to convert 5mC to cytosine. This enzymatic process depends on calcium ions and a reducing environment. The increased concentration of calcium ions in the early zygote (fertilized ovum) coincides with the period of active demethylation.
The important task for future research is to determine the exact role of different site-specific types of methylation via DNA methyltransferases in the normal genome and in the disease. Such information would be useful for prediction, diagnosis and treatment of diseases associated with abnormal DNA methylation.