Studying the activity of the genome, one cell at a time

DNA methylation

To understand how the human genome governs what a cell does and how it looks, it's important for scientists to understand what parts of the genome are activated in different cell types, and how these parts act together to make sure the cell functions properly. To study these questions, researchers have developed tools that measure the structure and activity of genetic material in cells.

In living cells the genome is tightly packaged in a structure called chromatin. Chromatin consists of the genomic DNA itself, proteins, and RNA. When the cell is actively transcribing and replicating DNA, the chromatin is more loosely packed or accessible. Scientists can measure chromatin accessibility to tell which regions of the genome are probably active at a given time. They can also measure levels of DNA methylation, when molecules are added to DNA to "turn off" parts of the genome, to see which regions of DNA might be repressed.

Sebastian Pott, a research assistant professor in the Department of Human Genetics at the University of Chicago, is working to answer some of these questions. In a new study published in the journal eLife, he tested a tool that measures both chromatin accessibility and DNA methylation to see if it could be applied to single cells. Science Life spoke to him about the work.

Why was it important to see if this method works for single cells?

Until recently, most of these studies were performed using large samples that contained thousands of cells. This is beneficial because the starting material is not limited. However, it has become clear that even cells that are seemingly of the same type differ in subtle but important ways. These differences are lost when using bulk samples, which are more likely to reflect the average of a cell population.

It is therefore necessary to develop methods that measure these features in single cells. Because the amount of starting material [from a single cell] is very small, it is critical to modify many of the established techniques when adapting them for single cell measurements. Studying features of the genome in single cells has the added limitation that the sample cannot be divided up for multiple assays. To address these limitations, this study was designed to test whether a method that measures both chromatin accessibility and DNA methylation, called NOMe-seq, could be applied to single cells. It builds on previous work by the labs of Peter Jones (from the Van Andel Research Institute) and Michael Kladde (from the University of Florida). In this study, I tested whether this protocol could be modified and applied to single cells; the modified protocol was named "single cell NOMe-seq," or scNOMe-seq.

Did it work?

This study provides evidence that scNOMe-seq detects endogenous DNA methylation and chromatin accessibility at the same time from a single cell. To test the performance of scNOMe-seq the study was performed in well-characterized human cell lines. Regulatory regions (e.g. promoters of active genes or active regulatory regions and enhancers) showed characteristic chromatin accessibility in single cells. As previously documented in bulk samples, DNA methylation in these regions was inversely correlated with chromatin accessibility. This proof-of-principle study therefore established that scNOMe-seq recovers the same features of chromatin organization observed in bulk samples.

What will scientists be able to do with a tool like this?

Many biological samples contain mixtures of cell types. As a result, it is often difficult when studying such a sample to extract a specific cell type for detailed characterization. For example, not all cell types in a tissue might be known in advance or have specific markers that would allow one to label them and separate them from the rest. In such cases, scNOMe-seq can be applied to measure regulatory activity (i.e. DNA methylation and chromatin accessibility) in individual cells without first isolating specific cell types. The resulting single cell data could then be used to classify and identify distinct cell types or states. This approach could be particularly useful when studying tumor samples which are often very heterogeneous.

What was your biggest challenge during this study?

Cells contain extremely small amounts of DNA and the biggest challenge of this work was to capture as much of the DNA in a cell to be able to measure anything at all. Only after very long process it is possible to assess whether this method works, and so it was pretty cool to see that this approach actually turned out data that were very similar to data produced from thousands of cells.

What are you working on next?

The aim of this study was to provide proof-of-principle that this method reliably obtains data from single cells. But of course, the motivation to develop the method was to use it to study biologically and clinically significant problems. For example, I am planning to apply this technique to study the activation of immune cells. I am particularly interested in understanding how responses of immune cells are regulated in individual cells and what might be different in cells of individuals with allergies.