DNA methylation has various roles in biology, including mediating gene expression. In embryonic development, DNA methylation acts as the epigenetic barrier that guides and restricts cell differentiation and bulk methylation as well plays a major role in silencing transposable and viral elements. In this way cells are not only guided towards intended lineages, but as well prevented from regressing into undifferentiated states.

It has been shown that methylation near gene promoters can vary widely depending on cell type, and higher levels of methylated promoters often correlate with little to no transcriptional activity. Different tissue types show varied levels of methylation, and many tumorigenic tissues may be observed with further elevated levels. Analysis and techniques used to study DNA methylation are therefore commonly used to study chromosomal patterns of DNA and/or histone modification.

Understanding the pathobiology of DNA methylation may also help in recognizing its roles in disease. Tumor suppressor genes have shown to be silenced in some cancer cells due to hypermethylation, which underscores its potential as a biomarker for cancers. Many maturation defects, like dendritic arborization or an impaired neuronal excitability, may be characterized by hypomethylated neurons providing evidence for the critical roles DNA methylation plays in neuronal maturation. Dysfunctional methylation processes have been linked with diseases related to cognitive ability, including Rett Syndrome, Fragile X Syndrome, Prader-Willi Syndrome and Angelman Syndrome.

Other neurological conditions, like hereditary sensory and autonomic neuropathy type 1 (HSAN1), are also related to DNA methylation dysfunction where patients develop dementia and hearing loss in adulthood. More recent studies have explored further the relation between DNA methylation and neurology and have found connections between impaired DNA methylation and mental disorders like schizophrenia and bipolar disorder.

Biochemically, DNA methylation is characterized by the addition of a methyl or hydroxy methyl group to the C5 position of cytosine to form 5-methylcytosine (5mC). Methylation is not usually a random occurrence, and will occur in phosphate linked cytosine-guanine dinucleotides (CpG) where the C position precedes the G. Methylation may also occur after being catalyzed by a group of DNA methyltransferase enzymes, as in DNA replication. Interestingly, CpGs in bulk DNA are often methylated, however methylation is mostly absent in CpG-rich regions, termed CpG islands, that are present near the 5' end of 50- 60% of all genes.

Though more commonplace in plant biology, non-CpG methylation can also occur in embryonic stem cells with CHH and CHG trinucleotides, where H can be any base nucleotide including adenine (A), cytosine (C), or thymine (T). Lysine residues in histone tails may also be susceptible to methylation under certain circumstances.

A number of techniques can be used if there is a need to evaluate DNA methylation on a genome-wide scale.

Semi-quantitative ELISA can provide a rough estimation of global DNA methylation, is often used for assessing larger scale changes in a genome, offers quick testing times and ease of use, and is commercially available in many kit formats. Experimentation follows general ELISA techniques; a primary antibody is selected against 5mC, a secondary antibody is labeled, colorimetric or fluorometric detection reagents are added, the sample is processed, and then the plate is read and analyzed appropriately.

Some ELISA kits specifically detect methylation levels of only long interspersed nuclear elements-1 (LINE-1) retrotransposons, of which roughly 17% of the human genome is composed, which are detectable at very low levels of DNA methylation (0.5%). Techniques that utilize high performance liquid chromatography (HPLC) may also be used due to their extremely high sensitivity.

HPLC-UV has quickly become the gold standard for quantifying deoxycytidine (dC) and 5mC in DNA samples. In HPLC-UV, the DNA is first hydrolyzed into its component bases, dC and 5mC bases are separated by chromatography, and the 5mC/dC ratio can be calculated for each sample. Ratios in test samples can be compared against test samples, to determine specific levels of methylation present.

LC-MS and/or MS alone do not require as much starting material as HPLC-UV, where typically 50-100 ng is suitable, and can detect methylation levels at extremely small levels, between 0.05 - 10%. LC-MS follows similar experimental steps as HPLC-UV, with the addition of mass spec that can measure the mass to charge ratio of molecules present in a sample, and is not adversely affected by poor-quality DNA derived from fresh-frozen paraffin-embedded (FFPE) samples. Though both techniques offer high sensitivity, the need for specialized laboratory equipment and experienced personnel may pose challenges.

The luminometric methylation assay (LUMA) is another useful technique that utilizes two DNA restriction digest experiments, performed at the same time. The first uses HpaII, a CpG methylation-sensitive enzyme, while the other uses MspI, a methylation-insensitive enzyme that cuts all CCGG sites.

Restriction Enzymes Cut Sites Reference Table
Restriction enzymes are a broad class of DNA-cutting enzymes that occur naturally in bacteria. Click the button on the right to check out our dataset table containing a full list of restriction enzymes.

Both techniques incorporate EcoRI as the internal control, and undergo subsequent pyrosequencing which fills in the overhangs left by enzymatic activity. The light produced can be measured and calculated as the HpaII/MspI ratio, which is linearly correlated to the amount of unmethylated DNA present in the test sample. This method offers high specificity, low variability, and requires small amounts of starting material, usually less than 500 ng, though it is crucial that starting DNA is of extremely high quality.

Levels of LINE-1 methylation may also be further explored by another series of techniques that involve:

Though this method is geared towards LINE-1 elements and is therefore not able to identify many CpG sites, it can reveal overall global DNA methylation changes accurately. Specifically, this method is useful for high throughput applications in assessing diseased or cancerous tissues where hypomethylation is apparent.

The most conventional method of DNA methylation analysis includes PCR-based amplification fragment length polymorphism (AFLP) and restriction fragment length polymorphism (RFLP), used separately or together. These methods offer quick experimental times, and general steps include DNA separation, restriction digest, then Southern blot. AFLP also includes an additional extraction step that aids in the analysis of the gel after electrophoresis.

These techniques come with limitations, and are less common than other DNA methylation analysis techniques. They can only assess a small percentage of global DNA methylation, and receiving good resolution of DNA bands may be tricky as a number of variables must be considered.

Other techniques are better suited for assessing specific genes and/or regulatory regions that are differentially methylated, as opposed to testing overall changes in methylation.

The most common technique is bisulfite sequencing, and unlike other sequencing technologies, it can easily distinguish 5mC from Cs present in a sample. Bisulfite treatment works by mediating the deamination of C to U, whose residues will be read as T after PCR and Sanger sequencing.

5mC residues, whose amino groups are protected, will however persist as C residues, and after sequencing all 5mCs can be acutely detected. Though bisulfite sequencing offers possible capabilities in conjunction with NGS to offer analysis across an entire genome, it does not come without challenges; bisulfite conversion reduces the genome complexity to only three base nucleotides making post-NGS alignment difficult, and bisulfite treatment leaves the sample DNA prone to fragmentation.

Types of Bisulfite Sequencing

Other restriction enzymes, including BisI, GlaI, KroI, PcsI, etc. are commercially available that use methylated DNA as a substrate, where some even function to cut outside of the recognition site. Such techniques require high quality DNA for digestion, though they have been helpful in the isolation of CpG islands. As these methods are bisulfite free, whole genome methylome profiling is also an option.