Epigenetics Experts Interview: Dr Michael Booth
Dr Michael Booth is a Postdoctoral Researcher in Professor Hagan Bayley’s lab and a Junior Research Fellow at Merton College, Oxford. During his PhD, which was carried out at the University of Cambridge under the supervision of Professor Sir Shankar Balasubramanian, Michael developed the technique of oxidative bisulfite (oxBS) sequencing which, for the first time, allowed single nucleotide resolution analysis of 5-hydroxymethylcytosine (5hmC). This invention has allowed much greater insight into the function of this important epigenetic mark, opening the door to potential prognostic, diagnostic and monitoring applications.
We spoke to Michael to find out more about the development of oxBS, his interest in epigenetics and his current research.
What first alerted you to epigenetics and the potential importance of studying 5hmC?
When I started my PhD in Professor Sir Shankar Balasubramanian’s group at the University of Cambridge, I mentioned that I had a big interest in epigenetics, which was a field that his group hadn’t really worked on at that time. Soon after, Shankar put Professor Anjana Rao and Professor Nathaniel Heintz’s 2009 Science papers , on my desk that showed the existence of 5hmC in mammals. Mass spectrometry, the main method of identifying 5hmC at this time, could only show the presence of 5hmC. So we set about trying to create a method that could accurately locate the modification in the genome. I came up with a couple of ideas, one being oxBS using oxidation and the other being a pull-down method that was subsequently developed and published by Dr Eun-Ang Raiber. The advantage of oxBS is that it has single nucleotide resolution, so I decided to put all of my efforts into that method.
The second publication was essentially an expanded and optimised version of the initial protocol. Dr Toby Ost had joined Shankar’s lab to spin-out the technology and he had done a lot of work improving the formulation of the oxidant, making it much more stable and reproducible. I had also been working on improving many of the steps – including the purification and PCR steps. The Nature Protocols format also allowed us to describe the methodology in much more detail.
In addition to creating the oxBS technique, you also applied it to genomic samples, could you tell us a little more about your findings?
That was the really important step. Professor Tony Green from the Department of Haematology at Cambridge University put us in contact with Professor Wolf Reik at the Babraham Institute. Wolf is one of the world leaders in epigenetics research and has published some incredible papers. Up until this point, I had done all the methodology and validation work on synthetic DNA, but through our collaboration with Wolf’s lab, we were able to access biological expertise and samples. Initially, working with Miguel Branco, who now has his own group at Queen Mary University of London, we used embryonic stem cells to see how the oxBS technique performed with genomic DNA. Fortunately, the technique worked very well without too many issues. Our work together identified some interesting genomic locations for 5hmC, including gene bodies and retrotransposons. Also, the technique allowed us to compare the locations of 5mC and 5hmC, which showed that in locations with intermediate levels of methylation, which had been previously thought to be plastic, there were high levels of hydroxymethylation, suggesting a link or interplay between the two.
What challenges did you face developing oxBS?
Easily the biggest challenge was finding an oxidant. It was quite a daunting task, as I had to find an oxidant that was specific for the primary alcohol in the presence of the other functionalities in DNA. Lots of oxidants don’t necessarily work in water and performing oxidation on the base itself was quite different to performing it on the base when part of a DNA strand. So there were many things to consider and I had to create lots of different assays to study the different oxidants. Upon screening all of the oxidants I could think of, or were suggested to me, that would work in water or part water, part organic solvent I found potassium perruthenate (KRuO4).
We saw some DNA damage from the oxidation, but bisulfite treatment itself is not a gentle technique, in that it cleaves the DNA a lot. Others have investigated alternative ways of sequencing methylation, but bisulfite sequencing is still the only method that really works well, hence being called the ‘gold standard’.
The analysis and interpretation of next generation sequencing data has become much easier over the last few years, but when you were creating oxBS, there would have been no specific analysis pipelines, so how did you approach this?
Analysis was really important. For the first part of the work on synthetic bases, I could do it all myself, but when we applied the technique to genomic DNA, we were really fortunate that Wolf Riek’s group really understood this area. We worked with Felix Krueger at the Babraham Institute who developed his own computational techniques to map methylation and on top of that, Miguel Branco, who was co-first author on the paper, was incredible. Miguel had the biological background and was learning the computational side and was able to bridge the knowledge gap between the teams.
How does oxBS compare with other methods of single nucleotide level 5hmC detection such as TAB-Seq?
I’ve not run TAB-Seq; however, the techniques are quite different as they have different positive outcomes. In TAB-Seq the positive result is hydroxymethylation, whereas in oxBS the positive result is methylation and then you have to compare with bisulfite sequenced DNA to identify the hydroxymethylation. However, while TAB-Seq directly assays the 5hmC, I don’t think there are many applications where you would just want to sequence 5hmC and not 5mC. One potential limitation of TAB-Seq is that it is an enzymatic rather than chemical technique and, as a result, the conversion efficiency is reliant on the activity of the enzyme. The enzyme may also exhibit some sequence specificity, thereby biasing the data to certain genomic regions. However, it may be possible to overcome these issues.
How important do you think it is to study 5hmC in addition to 5mC?
I know I am biased, but I think it is absolutely vital. Studies in brain cells show that 5hmC is up to 25% of the level of 5mC. If I was to review a bisulfite sequencing paper, unless they had done mass spec to show there was no 5hmC in the genome, then I don’t see how they can trust their data as only 5mC. As there are commercially available products that now allow 5hmC to be identified and located, why would you not test for it? However, I do understand that it adds to the expense of the research and this is probably the main factor limiting more widespread use.
I believe there needs to be more awareness that BS sequencing detects both 5mC and 5hmC, and so does not resolve them. So DNA methylation studies based on BS alone are potentially missing some important findings.
Your current research is no longer focused on oxBS or epigenetics but do you maintain an active interest in the field?
I try to read the papers – especially those from Shankar’s group and Wolf’s group to see what they are doing and where they are pushing different techniques, not just oxBS. There is a lot of excellent work now on other modifications from Chuan He’s group in Chicago, who invented TAB-Seq, a competitor to oxBS. They are working on methylation of RNA.
Where do you see the future of epigenetics?
Wolf Reik’s group at the Babraham Institute has done some excellent work recently on single cell bisulfite sequencing and I think this is a really exiting area for the future.
I believe the most important application of epigenetics of the discovery of new biomarkers. There are already drugs on the market that are targeting methylation, so I’d be really interested to know if you can target hydroxymethylation in the same way. Identifying 5hmC cancer biomarkers could expand the use of liquid biopsies to detect disease.
What new breakthroughs do you think are required to further drive epigenetics research?
One key barrier to adoption of these new epigenetics techniques, such as oxBS, is the lack of awareness that they work with existing analysis technologies. When I was investigating potential ways to sequence 5hmC, it was really important to me that the downstream analysis could be performed using existing platforms that users are already comfortable with. That is one of the reasons I pursued my idea of the oxBS technique, as it builds on the already well-utilised method of BS conversion. I also feel that as sequencing costs continue to decrease, epigenetics research will also expand significantly, as this is currently a key hurdle to larger studies.
For the development of oxBS, how important was it to be part of a multidisciplinary team, and are such teams becoming more commonplace in research?
When I joined Shankar’s group there was a little bit of biology going on, but not to the extent it is now. The collaboration with Wolf’s group was incredible and it certainly broadened my horizons. We had many full day meetings where we presented our work and this process was really important for developing our research and me as a scientist. It was interesting to see Shankar’s group change with this collaboration as he brought more biologists and bioinformaticians into the group.
Multidisciplinary science is a big focus at the moment. Where I am now, in Professor Hagan Bayley’s group at Oxford University, he has everyone from biologists, to physicists, to chemists and engineers — it’s a thoroughly varied environment. I’ve been lucky to work in groups with such broad expertise, so if ever you need anything, the person sitting next to you probably knows it. You can’t underestimate how important that is. Also, as a researcher in Cambridge or Oxford, we are very lucky that the collaborators we want to work with are often just over the road.
That moves us nicely on to your current research.
As I mentioned earlier, when I started my PhD I had a big interest in epigenetics. By the end of that time, I had a big interest in synthetic biology. I wanted to use some of the techniques that Hagan had already invented, but adding in some of my own stuff using my knowledge of the chemistry of DNA.
Hagan had been working on aqueous droplet networks, composed of water droplets in oil. The oil contains a lipid, so if you bring together two droplets, each with a lipid monolayer, you get a bilayer between the two – like on a cell surface. You can functionalise the bilayer with membrane proteins. You can also make them into interesting things like biobatteries or electrical components. Before I joined there was an incredible PhD student called Gabriel Viller who developed a 3D printer for these droplets, allowing 3 dimensional networks to be created comprising thousands of droplets in predefined locations. At that point it was very simple – everything inside them was either salt or low concentration purified proteins. I wanted to put a lot more complexity into it by making them express protein and put them together to form what I call ‘synthetic tissues’. On top of this, I wanted to make the DNA inside the synthetic cells light activated. So I have been using my knowledge of the chemistry of DNA to chemically block promoters, so there is no protein expression until light is shone on the DNA. I have published this work now in Science Advances.
What potential applications does this technology have?
This is perhaps more basic research than I was doing in Cambridge. These aqueous droplets are a completely new type of material. The idea would be to create synthetic tissues that can communicate with living tissues. In the Science Advances paper, I showed it was possible to get electrical communication to go through the synthetic tissue, so potentially we could create a synthetic nerve. Importantly, as it isn’t a real cell, you wouldn’t get the immune reaction or rejection. On top of that you could use them as drug delivery devices to target specific cells. The facility for light activation allows drug delivery in a much more targeted way. Because it is such a new material, I am currently trying to find its limits to work out which of the potential applications are more likely to work.
Where can we hear more about your research?
I am currently looking at Synthetic Biology conferences for this year, but I recently presented my work at the 3rd Synthetic Biology Congress in London, a Pharmacy and Pharmaceutical Sciences seminar at Cardiff University and the Pint of Science event in Oxford. You can also look out for my new papers! Readers can also follow me on twitter @DrMichaelJBooth.
- Light-activated communication in synthetic tissues.Sci Adv. 2016, 2(4)
- Chemical methods for decoding cytosine modifications in DNA. Chem Rev. 2015, 115(6)
- Quantitative sequencing of 5-formylcytosine in DNA at single-base resolution. Nat Chem. 2014, 6(5)
- A screen for hydroxymethylcytosine and formylcytosine binding proteins suggests functions in transcription and chromatin regulation. Genome Biol. 2013, 14(10)
- Oxidative bisulfite sequencing of 5-methylcytosine and 5-hydroxymethylcytosine. Nature Protocols 2013, 8
- Quantitative Sequencing of 5-Methylcytosine and 5-Hydroxymethylcytosine at single-base resolution. Science 2012, 336(6083)
Find out how CEGX can support your epigenetics biomarker development and research projects.
- Tahiliani et al (2009) Conversion of 5-Methylcytosine to 5-Hydroxymethylcytosine in Mammalian DNA by MLL Partner TET1. Science 324(5929): 930–935
- Kriaucionis and Heintz (2009) The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain. Science. 324(5929):929-30