Differential DNA methylation analysis optimally requires purified cell populations

Michael K Skinner

doi:10.1016/j.fertnstert.2016.06.008

Because each cell type has a unique epigenome to help provide cell specificity, the use of a mixed cell population in the analysis of DNA methylation is a reflection of all cell types present. Therefore, changes in cell population number or composition can suggest differential DNA methylation that is not due to alterations in DNA methylation. The analysis of differential DNA methylation optimally requires the use of purified cell populations. There are over 200 cell types in the human body, and each tissue contains multiple cell types. Blood contains over 20 different cell types. Although all cell types contain the same DNA with a similar DNA sequence, each cell type has a very distinct epigenome to give the cell its cell specificity. Therefore, the difference between a neuron and a hepatocyte is in large part due to the distinct epigenomes of the two cell types. Because all cell types have the same DNA and sequence, genetic studies can easily use mixed cell populations as well as the whole organism, if required, to do genetic analysis. In contrast, epigenetic investigations, such as differential DNA methylation analysis, require consideration of mixed cell populations and purity of the cell types used. One of the first types of studies to investigate environmental epigenetics in humans used twin studies with discordant disease and whole blood analysis (1). More recently the need for the use of purified cell types is appreciated in epigenome-wide association studies EWAS) with the use of purified cell types (2). Although interesting observations have been provided with blood analysis, due to the presence of over 20 cell types interpretation of the observations is difficult. Small changes of 10% within different cell populations can generate a change in DNA methylation that is not due to alterations in DNA methylation at the specific genomic sites. Subsequent analysis of purified cell types would be required to confirm the differential DNA methylation analysis. There have been attempts to develop bioinformatics approaches to assess epigenetics in mixed cell populations (3), but without extensive information on different cell types epigenomes this type of approach is in its early days. When simply mapping epigenetic marks to the genome and not looking at differential DNA methylation regions (DMR), this issue of mixed cell populations is less problematic. There are also specific epigenetic sites that are consistent between cell populations, and a current example of this is imprinted You can discuss this article with its authors and with other ASRM members at

Differential DNA methylation analysis optimally requires purified cell populations

Files and links (2)

Abstract

Metrics

Details