May 31, 2018

Story and photo by Nadir Balba

The Paper of the Month for May 2018 is "Highly scalable generation of DNA methylation profiles in single genes" published in Nature Biotechnology. This project was led by the lab of Andrew Adey, Ph.D., assistant professor of molecular and medical genetics, OHSU School of Medicine.

Its authors in descending order are, Ryan Mulqueen, OHSU Department of Molecular and Medical Genetics, Dr. Dmitry Pokholokan and Steven Norberg,  Illumina Inc., Kristof Torkenczy and Andrew Fields, OHSU Department of Molecular and Medical Genetics, Dr. Duan Chen Sun, OHSU Department of Molecular Microbiology and Immunology, Dr. John Sinnamon, Vollum Institute, Drs. Jay Shendure and Cole Trapnell, University of Washington Department of Genomic Sciences, Dr. Brian O'Roak, OHSU Department of Molecular and Medical Genetics, Dr. Zheng Xia, OHSU Department of Molecular Microbiology and Immunology, Dr. Frank Steemers, Illumina Inc., and Dr. Adey.

A daunting task

One of the most fundamental challenges in basic science research is the problem of cell heterogeneity. There are over 35 trillion cells within the human body, and each cell can vary in several factors, including its environment, and which segments of DNA are actively being transcribed. These distinct cellular phenotypic profiles complicate biological research and make designing useful treatments even more challenging.

Quickly and efficiently distinguishing cell subtypes would be enormously beneficial, to both researchers and clinicians, by aiding basic scientific research and promoting more effective treatments for patients. Although it may seem like a daunting task to catalog them all, researchers at OHSU have taken some big steps towards this lofty goal.

Analyzing DNA methylation

There is roughly three feet of DNA packed into every cell, but only specific sections within this DNA sequence are transcribed, eventually leading to protein synthesis. One way cells accomplish this type of specificity is through DNA methylation, a chemical modification which modulates the transcription of a DNA segment without any changes to the DNA sequence itself. During this process, much of the DNA is coated with a variety of chemical markers that dictate which regions of DNA should be used and which should be ignored.

DNA methylation has been studied for decades, but previous research involving these markers has relied on quantifying large populations of cells all at once, generating an averaged profile of methylation across thousands to millions of cells. However, tissues within the body are highly complex and have a variety of different cell types hidden within them. For instance, tumors contain several different types of cells, each with a different pattern of DNA methylation. The same goes for cortical tissue, which contain both neuronal and non-neuronal cells. Hence, there is a fatal flaw in relying on these bulk-averaged methods due to their inability to deal with this inherent cell heterogeneity issue.

Over the past six years, other researchers have developed single-cell technologies to address this issue, but these methods lacked the necessary high-cell count, making them less efficient. They were simply too slow and too expensive to implement while studying complex cell populations, such as brain tissue or cancerous tumors.

Pioneering a more effective method

Dr. Adey's lab has been spearheading different ways to increase the cell counts of these methods. The team recently collaborated with several other researchers here at OHSU, as well as ones from the University of Washington and the San Diego-based biotechnology company Illumnia, to develop a procedure called single-cell combinatorial indexing for methylation, or sci-MET for short. This new technique can analyze DNA methylation markers in thousands, to tens of thousands of cells, simultaneously.

Although the team was hopeful for their newly developed technology, they had to prove it could work first. They were able to validate the technique by comparing its output against well-established signals, from both the bulk and single-cell methods, and found no major differences. Furthermore, they were also able to successfully measure methylation and discriminate cell types in a mixture of different cultured cell lines. They repeated the process using harvested mouse cortical tissue and were able to identify specific populations of neurons and non-neuronal cells, providing further evidence that the new process was just as precise as the previously established methods.

Faster and cheaper

But what's really impressive was how efficient this new process was. By using the sci-MET technique, the team was able to analyze over 3,000 cells at once, more than a 20-fold increase compared to the previously existing method. Even more astounding was how they were able to decrease the cost to prepare the cells for sequencing, from $20 to $50 a cell to less than 50 cents a cell, more than a 70-fold decrease in price.

Mary Heinricher, Ph.D., associate dean for basic research, OHSU School of Medicine, first heard about sci-MET from her OHSU colleagues who were excited about using the technology in their own experiments. "I knew I had to pick this paper after two people told me they needed to talk to Andrew about putting this cool new technology to work in their own system."

Looking to the future

With the method now developed, Dr. Adey's lab is looking forward to answering important biological questions about the propagation of DNA methylation in different diseased states, including which cell types are affected and where does DNA methylation begin to deviate during the pathology.

"One major motivator in this work is the ability to interrogate complex tissues for this important mark," said Dr. Adey. "Tumors across many forms of cancer have shown defects in regulating DNA methylation, which could lead to hints of how we can develop better diagnostics and interventions."

In addition, this method could be used to elucidate the processes affecting the brain during development. Many neurodevelopmental disorders show significant changes to DNA methylation across neuronal cell types. Prior to the development of sci-MET, analyses of where, when and on what cell types these disorders were altering had been relatively intractable. This technique will give scientists the ability to explore new avenues of research and will hopefully provide more precise treatments for clinicians in the near future.

Pictured above, from left to right: Dr. Andrew Adey, Kristof Torkenczy, Dr. Brian O'Roak, Andrew Fields, and Ryan Mulqueen.

Resources

• Read the paper

Citation

Mulqueen RM, Pokholok D, Norberg SJ, Torkenczy KA, Fields AJ, Sun D, Sinnamon JR, Shendure J, Trapnell C, O'Roak BJ, Xia Z. Highly scalable generation of DNA methylation profiles in single cells. Nature biotechnology. 2018 May;36(5):428.

More Published Papers

• Previous School of Medicine Papers of the Month
• Recent OHSU published papers