The Role of DNA Methylation in Carcinogenesis

With the sequencing of the human genome, a great deal of attention has been paid to various sequence polymorphisms and mutations, and which of these are likely to be linked to various disease states. However, the maintenance of normal patterns of genetic expression is a multidimensional process, and DNA sequence alone is only one of those dimensions. Another, often overlooked, aspect involves epigenetic mechanisms, which regulate genetic expression without altering the DNA sequence itself.

DNA methylation is an example of an epigenetic mechanism, whereby gene transcription is controlled through the addition of methyl chemical groups to cytosine bases at CpG dinucleotides. The methylated cytosines inhibit transcription factor binding and contribute to a more condensed chromatin formation. This can work in one's favor if an increased amount of methylation decreases the expression of a gene that contributes to obesity, but it can be detrimental if an increased amount of methylation contributes to the silencing of a tumor suppressor gene. Thus, both too much and too little DNA methylation can lead to adverse effects. DNA methylation normally silences transposable elements and viruses throughout the genome that could potentially cause harm if they were expressed. At the same time, CpG-rich regions of DNA in the promoter regions of tumor suppressor genes known as "CpG islands" are typically unmethylated, facilitating expression of these genes.

Aberrant patterns of DNA methylation leading to altered states of gene transcription have been implicated in numerous developmental, neurological and metabolic disorders including, but not limited to, Beckwith Wiedemann syndrome, autism and diabetes. In addition, altered DNA methylation is tightly linked to cancer. Typically, the level of DNA methylation at CpG sites of viruses and transposable elements is decreased in neoplastic tissues, while the level of methylation at various promoter region CpG sites is typically increased. The overall effect is a "worst of both worlds" situation in which the expression of potentially harmful transposable elements is increased, and the expression of tumor suppressor genes that are frequently controlled by CpG island methylation is decreased. Little is understood about the mechanisms that lead to increased methylation in some areas of DNA and concurrent decreases in others. However, there is a good deal of similarity in the patterns that are altered in neoplastic tissue. An enhanced understanding of DNA methylation changes throughout carcinogenesis and what influences these changes can lead to the development of useful biomarkers as well as chemopreventive strategies.

Diet plays an important role in DNA methylation. The availability of "methyl group donor" nutrients and enzymes in the one-carbon metabolic pathway is necessary for the maintenance of DNA methylation. Choline, methionine, zinc, B-12, betaine and folic acid are all key nutrients in this cycle and are obtained from dietary sources, such as leafy green vegetables, eggs and meats. Many studies have indicated that deficiencies in nutrients involved in the one-carbon pathway are harmful. Clinical studies have shown that folic acid deficiency has been linked to an increased risk of colon cancer, neural tube defects and cardiovascular disease. Since either too much or too little DNA methylation can have an adverse effect, it is logical to assume that excessive amounts of these nutrients can also be harmful. There have been recent animal and epidemiological studies that suggest large amounts of folic acid increases the risk of breast cancer, although additional research is needed to fully substantiate this claim.

	"A choline and/or methionine deficient diet leads to an overall decrease in DNA methylation, increased expression of transposable elements, and an increase in spontaneous hepatocarcinoma in animal studies." Rebecca Watson, PhD Research Biologist

At IITRI, we are interested in further characterizing the relationship between diet, DNA methylation and cancer risk. Such research could help distinguish between beneficial and harmful amounts of various nutritional supplements and chemopreventive agents that influence DNA methylation patterns. Additionally, the role of maternal diet in the methylation patterns of offspring is particularly intriguing. During development, DNA methylation patterns are first being established, and thus, the developing embryo might be especially sensitive to dietary components. Studies in rodents have shown that maternal diet can play a significant role in establishing long-lasting patterns of DNA methylation that can impact gene expression and perhaps susceptibility to disease as an adult. Given that folic acid is added to the food supply to decrease the incidence of neural tube defects, and the frequent recommendation that pregnant women take folic acid supplements, an enhanced understanding of this relationship is of great importance to public health.

Another potential avenue of research at IITRI is detecting specific DNA methylation alterations for use as biomarkers of cancer risk. This would be useful both in the clinic as well as in the development of improved animal models for carcinogenesis. Recent work at Johns Hopkins has shown that DNA methylation status of the insulin-like growth factor gene 2 can identify people at high risk for colon cancer. The ideal biomarker in an animal model would be one that reliably predicts cancer susceptibility in an area of the genome in which parallel changes are seen in humans, and can be detected in the serum of the live animal. Given that similar methylation changes are often seen in rodents and humans in the same types of cancer tissues, DNA methylation changes that are found to reliably predict tumor formation would be extremely useful in the development of cancer models for use in determining the carcinogenic and/or chemopreventive potential for various substances. The development of such a biomarker could result in a significant savings in time and money, and would greatly facilitate cancer research, particularly for cancer types that involve a long latency period.

DNA methylation alterations are found in nearly all neoplastic tissues, and are an important part and/or consequence of the carcinogenic process. Further characterization of the role of DNA methylation in cancer has the potential to provide results that could improve our understanding of the mechanistic underpinnings of cancer in a manner that can have a real impact on human health.