Nature News recently described a trend noticeable at Emory and elsewhere. That trend is epigenomics: studying the patterns of chemical groups that adorn DNA sequences and influence their activity. Often this means taking a comprehensive genome-wide look at the patterns of DNA methylation.
DNA methylation is a chemical modification analogous to punctuation or a highlighter or censorâ€™s pen. It doesnâ€™t change the letters of the DNA but it does change how that information is received.
One recent example of epigenomics from Emory is a collaboration between psychiatrist Andrew Miller and oncologist Mylin Torres, examining the long-lasting marks left by chemotherapy in the blood cells of breast cancer patients.
Their co-author Alicia Smith, who specializes in the intersection of psychiatry and genetics, reports “EWAS or epigenome-wise association studies are being used in complex disease research to suggest genes that may be involved in etiology or symptoms.Â They’re used in medication or diet studies to demonstrate efficacy or suggest side effects.Â Â Theyâ€™re also used in longitudinal studies to see if particular exposures or characteristics (i.e. low birthweight) have long-term consequences.”Other recent examples include:
Epigenetic changes in heart failure, from pathologist William Lewisâ€™ lab
Smoking-related epigenetic changes compared in African Americans and Caucasians, from Rollins epidemiologist Yan Sun and colleagues
EWAS follows in the footsteps of genome-wide association studies or GWAS. There was a boom in GWAS about a decade ago, driving the discovery of genetic variations linked to many diseases. But geneticists have also noted limitations in GWAS studies: they generally look at common variants (single nucleotide polymorphisms) and thus donâ€™t necessarily capture all of whatâ€™s responsible for complex diseases such as autism or schizophrenia, for example.
Similarly, EWAS faces particular challenges. In the human body, all of an individualâ€™s DNA is the same. [Well, considering that there are random somatic mutations sometimes leading to tumors and white blood cells that rearrange their DNA on purpose, scratch that. But to a first approximation…] For epigenomics, the differences between different cell types make things even more complicated.
And not everything that looks like methyl-C on bisulfite sequencing is the same. Geneticist Peng Jinâ€™s team has been probing the complexity that underlies the phenomenon of hydroxymethylation; sometimes a signal that usually means â€œthis gene is offâ€ actually means â€œready to be turned on.â€
We asked Paula Vertino, who studies epigenetic effects in cancer, to summarize some of the other complexities with EWAS:
In my mind the biggest challenges with EWAS studies is not in the technology; the platforms available to measure DNA methylation genome-wide are robust, but rather in the interpretation.
We actually know surprising little about how DNA methylation at sites outside of gene promoters affect gene expression. This is akin to the challenges encountered in the early GWAS studies, where many disease-associated SNPs are found outside of coding regions, and thus translating a change in DNA sequence to a specific biologic function can be difficult.
Another consideration particular to EWAS is that unlike GWAS studies where the disease or trait is linked to a difference in DNA sequence, and is a reflection of germline variation so the pattern is constant in every cell in the body, for EWAS studies, DNA methylation differences are measured in shades of gray, and are a reflection of the mixture of cells in the starting material.
In GWAS the assumption is that you are born with that genetic variant and it in some way drives that trait. For EWAS, itâ€™s not so simple because you are looking at a methylation pattern of the particular somatic tissue that you are analyzing (often blood cells), which is not necessarily reflective of the epigenome you were born with nor of the cells or tissues that underlie the disease or phenotype. Â