The Role of DNA Methylation in Human Disease

DNA methylation is one of the earliest epigenetic modifications to be discovered in human beings. It involves the transfer of methyl (CH₃) groups to the C5 position of cytosine bases that comprise deoxyribonucleic acid (DNA) to produce 5-methylcytosine (5mC) – the reaction is catalyzed by a family of enzymes called DNA methyltransferases (DNMTs). Typically, the altered cytosine bases reside immediately adjacent to guanine bases. This leads to two 5mC bases sitting diagonally to each other on complementary DNA strands.

DNMTs have several distinct roles, for instance, they may function as de novo DNMTs, which involves establishing the initial pattern of methyl groups on a DNA molecule. While other DNMTs adopt maintenance roles, copying the methylation pattern from an existing DNA strand to its new partner after replication has occurred.

Several studies in the 1980s revealed that DNA methylation played a major part in both gene regulation and cell differentiation. Since then, further research has confirmed the role of abnormal methylation in the development and progression of various diseases. According to Manel Esteller, director of the Josep Carreras Leukaemia Research Institute and professor of genetics at the University of Barcelona, “DNA methylation is one of the main controllers for specific-tissue expression allowing the correct expression of a gene in the right organ or cell type.” He further added, “DNA methylation acts as a buffer to stabilize our genome and silence repetitive chromosomic regions. Many diseases show an alteration of DNA methylation that disrupts cellular activity.” Esteller’s research mainly focuses on alterations in DNA methylation, histone modifications and chromatin in human cancer. At present, he is working on establishing epigenome and epitranscriptome maps for normal and transformed cells.

In mammals, methylation is mostly sparse but is globally distributed in specific CpG or CG (cytosine–guanine) sequences. In certain regions of the genome, CpG is abundantly found (e.g., CpG islands). In healthy cells, CpG islands associated with gene promotors are typically free from methylation, whereas islands found within gene bodies tend to become methylated during development. Researchers have pointed out that methylation of CpG islands at promotor regions can cause inappropriate downregulation of specific genes (e.g., silencing of tumor suppressor genes in cancer cells). 

The role of DNA methylation in normal biologic processes 

The role and position of DNA methylation varies among different kingdoms of organisms. As mentioned above, mammals tend to possess a fairly global distribution of CpG methylation, while invertebrate animals typically show a "mosaic" pattern of methylation.

DNA methylation plays an important role in many biological processes, for example, genomic imprinting, stem cell differentiation and chromosomal stability, and is considered an essential modification that regulates cell growth and proliferation. DNA methylation patterns are mutable and inheritable and in the case of abnormal DNA methylation in the parental allele, various serious diseases, such as cancer, aging disorders, metabolic ailments, psychological disorders and genetic diseases, may occur.

DNA methylation and disease

Scientists first discovered the role of DNA methylation in human diseases while studying genomic imprinting. Genomic imprinting is a stable and heritable phenomenon that occurs independently from classical Mendelian processes. It involves the epigenetic marking of a gene (e.g., methylation), based on its parental origin which results in differential expression of the gene without modifying the underlying DNA sequence. Here we highlight some of the diseases caused by aberrant DNA methylation.

Autoimmune diseases     

Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease that causes symmetrical polyarthritis in small and large joints. Scientists performed genome-wide DNA methylation analysis of peripheral blood mononuclear cells and found altered DNA methylation of human leukocyte antigen (HLA) class II. This abnormal DNA methylation can facilitate the genetic risk of developing RA.

Systemic lupus erythematosus (SLE) is an autoimmune disease in which the body’s immune system incorrectly attacks its own healthy tissue. A genome-wide assessment of DNA methylation demonstrated differential DNA methylation in the genes of SLE patients, associated with autoantibody production. Abnormal DNA methylation was observed in the promoter region of the IL-6 gene.

Metabolic disorders         

Scientists have also linked conditions such as hyperglycemia and hyperlipidemia with differential DNA methylation, which leads to abnormal gene expression. Reduced methylation of cyclin-dependent kinase inhibitor 1A and phosphodiesterase 7B promoters revealed dysregulated glucose-stimulated insulin secretion in humans. DNA methylation has also been linked with obesity, where increased DNA methylation of hypoxia-inducible factor 3 alpha in adipose tissue and blood cells causes an increase in body mass index (BMI).

Cancer

Methylation is one of the earliest changes in many cancer types, and detecting the methylation state in cell-free DNA is a promising route to non-invasive early cancer detection. Recently, omics (genomics, proteomics, transcriptomics and metabolomics) studies have provided supporting evidence of these findings, confirming that aberrant DNA methylation is linked to hepatocellular carcinoma, breast cancer, glioblastoma, squamous cell lung cancer, leukemia and thyroid carcinoma. Differential levels of expression of DNMTs and mutations in DNMTs are commonly found in cancer patients, both of which affect methylation mechanisms. Previous research has shown that epigenetic dysregulation plays a crucial role in tumor development and metastasis.

“In cancer, we observe a general global DNA hypomethylation of the genome and more focal DNA hypermethylation that affects CpG-rich sequences (so-called CpG islands) often found at the promoter,” explained Professor Gerd Pfeifer, from the Center for Epigenetics, Van Andel Institute. Pfeifer’s laboratory investigates the underlying mechanisms of cancer and other diseases, specifically focusing on DNA mutations, DNA methylation and the role of 5mC oxidation. According to Pfeifer, most of the DNA hypermethylation events in cancer are inconsequential because the genes are already silent. However, some methylation events can be considered tumor drivers, when, for example, they silence genes encoding anti-proliferative factors, DNA repair genes, or genes essential for normal cell differentiation.

Paula Esteller-Cucala is a doctoral researcher in the Comparative Genomics Group at the Institut de Biologia Evolutiva (IBE), her work focuses on epigenetics and transcriptomics of non-human primates. She said, “Methylation patterns are very heterogeneous. They might differ from one cancer type to another and also from one cell type to another cell type. Understanding the role of these modifications and their effect in different cancer types is essential to target potential treatments and therapies.” Identifying cancer-specific DNA methylation markers (regions of the genome that are specifically methylated or unmethylated in one or more cancer types or subtypes), can be used to detect and monitor cancer with a view to developing therapeutic strategies.

Gliomas are a common type of brain cancer, which originate in the glial cells that support neurons in the brain. Recently, researchers have used a single-cell multiomics approach to identify methylation marks within individual tumor cells obtained from patients with glioma. They were able to confirm distinct patterns of DNA methylation responsible for shifting the cells from one state to another (e.g., stem-cell-like states to mature states) and developed a map of cell states from the sampled tumors. The insights gained from the study could help to develop better ways to detect, stage, monitor and treat the disease.

Neurological disorders   

Aberrant DNA methylation is also associated with neurological disorders. Mutation in the methyl-CpG-binding protein 2 gene has been found in patients with Rett syndrome and autism spectrum disorder. Scientists found that changes in the DNA methylation profile promoted altered expression of genes associated with synaptic activity.

A lowered methylation level of catechol-O-methyl transferase in peripheral blood was observed in patients with schizophrenia.         

An epigenome-wide association study compared the methylation patterns of tissues from three different mammalian species to determine if Huntington’s disease is accompanied by altered DNA methylation. The researchers found that the disease was associated with “profound changes” to the level of DNA methylation.

A systemic review of DNA methylation in Alzheimer’s disease found that the APP gene – encoding a protein called amyloid precursor protein which has been associated with the formation of amyloid plaques – is consistently hypermethylated in brain and peripheral blood.

Technologies used to investigate DNA methylation

According to Pfeifer, “Most of the conventional methods for the analysis of DNA methylation patterns are based on converting cytosine to uracil by chemical deamination using high temperature and high concentrations of sodium bisulfite, while some newer methods use enzymes for deamination.” Typically, pathologists face difficulties in predicting biomarkers for cancer and representing the overall status of tumor mass from small biopsy tissue samples or liquid samples obtained from patients with lung cancer. In this context, Pfeifer, stated that “these new enzymatic approaches could allow researchers to work with very small amounts of DNA (biopsy samples).”

Looking beyond traditional methods, recent advances in sequencing and array technologies have enabled researchers to conduct detailed DNA methylation profiling, providing a comprehensive picture of its role in disease. In Esteller-Cucala’s opinion, the latest methodology used to study DNA methylation is by means of long-read sequencing – these technologies allow much longer sequences to be read (> 10000 bp).

Esteller, also provided his thoughts, “The technology most widely used to study, in a cost-effective manner, human DNA methylation is based on DNA methylation microarrays that interrogate 850K CpG sites of our genome.”

Some techniques used to determine DNA methylation are discussed in more detail below.

Bisulfite-based assays

Bisulfite treatment is a commonly used method, introduced by Frommer et al. (1992), for analyzing 5mC and nonmethylated bases. In this method, the genomic DNA is exposed to sodium bisulfite which promotes nonmethylated cytosine deamination and converts it to uracil, while the methylated cytosine remains unchanged. Finally, subsequent PCR amplification converts the uracils to thymines. This results in the transfer of gene methylation information to the newly formed sequencing libraries. Scientists use the bisulfite sequencing method to analyze 5mC in the resulting single strands. This technique can be used to determine multiple DNA methylation events; however, multiple steps such as cloning and sequencing processes make it a time-consuming method. Some of the other limitations of this method include degradation of DNA samples owing to the aggressive chemical conditions required, such as low pH and high temperature to achieve deamination. Hence, sample degradation leads to a decrease in sensitivity, making it unfit for analysis of samples that are low in concentration, e.g., ctDNA. However, less aggressive treatments fail to convert all unmethylated cytosines and, thereby, may overestimate the levels of methylation.

An advanced sequencing-based technique known as methylation-specific PCR (MS-PCR) has been developed which avoids the complex sequencing process.

Enzymatic techniques

Detection of methylation based on enzymatic methods is quick with high specificity under mild reaction conditions, meaning it can achieve the same end product as the bisulfite method, without compromising the integrity of the DNA. Enzymatic methyl-sequencing (EM-seq) can detect 5mC and 5-hydroxymethylcytosine (5hmC) using three enzymes, namely, T4-phage beta-glucosyltransferase (T4-BGT), Tet methylcytosine dioxygenase 2 (TET2) and apolipoprotein B mRNA editing enzyme catalytic subunit 3A (APOBEC3A). EM-seq is a two-step reaction process. The first step involves glucosylation of 5hmC via TET2 and T4-BGT enzymes which develops products that cannot be deaminated by APOBEC3A. The second step involves the APOBEC3A enzyme, which deaminates unmodified cytosines by converting them to uracils.     

Some examples of methylation-sensitive restriction enzymes (MREs) include HpaII, BstUI, NotI and SmaI. These enzymes only cut the nonmethylated target regions and keep the methylated DNA intact. These MRE cuttings are subsequently sequenced to predict the DNA methylation levels at the genomic level. Recently, scientists have developed an advanced enzymatic digestion technique, called methylation-sensitive restriction endonuclease-PCR/southern (MS-RE-PCR).

Direct oxidation and chemical decomposition of oxidation

Scientists have developed an electrochemical-based method for the direct analysis of DNA methylation. For this technique, choline chloride monolayer-supported multi-walled carbon nanotubes (MWCNTs) (MWCNTs/Ch/GCE) were designed. This technique is highly specific, precise and rapid and does not require the use of enzymes, probes, or bisulfite. Another new chemical method has been developed, which is based on chemical oxidation decomposition, which can separate 5mC. This method uses light-sensitive oxidation using the 2-methyl-1,4-naphthoquinone-chromophore. The chemical cut-off method efficiently detects the methylation sites.

Methodological challenges in DNA methylation studies

Many methods are used to study DNA methylation, and each has its own limitations. For instance, as stated above, the chemical treatment associated with the bisulfite-based method causes DNA degradation and shorter DNA fragments are obtained for further analysis. While long-read sequencing technology is considered a highly accurate approach, Esteller-Cucala pointed out, “the main caveat of this method is that you need good tools (algorithms) in order to detect the methylated positions”.

Pfeifer pointed out some additional considerations, "One challenge is to achieve coverage of the whole mammalian genome, which has over 25 million CpG sequences that can be methylated. To perform a quantitative analysis of the methylation state of each one of these CpGs, deep sequencing coverage is required, which is still expensive. There are more affordable methods available that can be used to analyze subsets of CpGs, but these methods may miss some critical methylation changes.”

For comprehensive DNA methylation studies, a large amount of DNA may be required and therefore, analysis becomes challenging when the tissue samples are scarce. “Sometimes it is difficult to distinguish 5mC from 5hmC,” said Esteller.

DNA methylation and clinical benefits

Abnormal DNA methylation is associated with many diseases and DNA methylation-based biomarkers can help to improve disease prognosis and treatment response. “One area of clinical interest in the DNA methylation field is to exploit altered DNA methylation patterns for cancer diagnosis. These alterations can be detected in “liquid biopsies,” derived from patients’ blood samples. Ultimately, early detection of cancer, based on changed methylation profiles in cell-free plasma DNA, could become possible,” explained Pfeifer. He continues, “It would be of interest to know if a methylation change present in a benign or cancer precursor lesion has the potential to promote progression to a malignant state.” Several companies are currently working to develop early cancer detection tests based on methylation state and AI-based prediction of pathology.  

Esteller explained that from knowledge of the DNA methylation landscape of tumoral cells, three translation uses have emerged in the oncology field: “The discovery of new biomarkers of the disease that can even be detected in biological fluids and allow its pathological classification; the use of DNA hypermethylation events in certain genes as predictors of response to therapies, helping cancer precision medicine; and the use of DNA methylation as a target for epigenetic drugs such as inhibitors of DNA methylation that are being used in the treatment of hematological malignancies.”