Epitranscriptomics – An Interview with Samie Jaffrey, Professor of Pharmacology, Weill Cornell Medicine
The study of epitranscriptomics is a fascinating and important one, with ramifications for better understanding disease symptoms and pushing forward drug treatment in a number of areas. But it’s also an emerging area still misunderstood by many and held up by certain controversies that have divided expert opinion for years. Front Line Genomics spoke to Dr Samie Jaffrey, Professor of Pharmacology, Weill Cornell Medicine, about the field.
FLG: Could you Provide a Broad Overview of Epitranscriptomics, and What it Means for Gene Expression?
SJ: The idea that nucleotides in mRNA can get modifications was discovered way back in the ‘70s, when people discovered that internal RNA nucleotides, in particular adenosine, can sometimes get a methyl modification. Scientists determined it was on the base portion of adenosine at the number 6 position, and so named it m6A.
Jim Darnell at Rockefeller University, one of the pioneers of molecular biology, first identified its function in 1978: RNAs with m6A seem to degrade more quickly than others. Therefore the major function of m6A is to control the stability of transcripts in a cell. Gene expression centres around one thing: controlling the amount of mRNA you have for every individual gene in a cell. You can control the amount of mRNA by either controlling how much RNA is transcribed at every gene or how quickly the RNA degrades. m6A is one way to control the degradation rate.
Because it was so difficult to detect and good tools were not available, m6A was forgotten about until recently. This happened largely because the key tool in molecular biology is taking mRNA and converting it to DNA through reverse transcription. This makes complimentary DNA, or cDNA. But m6A and adenosine both reverse transcribe to T, thymidine, meaning when you make your cDNA you lose any record of the m6A that’s in your transcript and where it’s located. People were reverse transcribing and sequencing their genes without ever knowing there was m6A present.
Then a 2008 study by Rupert Fray knocked out the enzymes that make m6A in plants and yeast. As opposed to killing the cells, it affected very specific pathways involved in cell fate and development. When we read that paper, we realised m6A may be a major regulatory mechanism. We immediately started working on a method to determine which transcripts contain m6A. The result of this work was our first mapping study in 2012, in which we developed an approach where we could use antibodies against m6A to immunoprecipitate mRNA fragments which have m6A in them. This approach showed us that m6A is only on very specific transcripts which were often functionally linked to cell fate and development, as predicted by Fray.
This method allowed anyone to determine which transcripts are getting methylated in any cell type they are studying. If the pattern of m6A in cellular mRNA is perturbed, researchers can identify transcripts that are no longer methylated properly, and explore how this relates to disease formation.
There’s now been over 500 m6A mapping datasets that have linked m6A changes and dynamics to diverse cancers, cell types, disease processes and drug treatments. This has allowed people to understand what sort of transcripts are being regulated by m6A, and how this is dynamically controlled in different cellular conditions. As well as degradation, m6A could affect RNA localisation or splicing to change the fate of the mRNA in the cell.
FLG: In the Past There Have Been Misunderstandings Around the Field’s Name. Could you Talk Us Through That?
SJ: The term “RNA epigenetics” itself is a controversial term. The preferred term is “epitranscriptomics”. In the past there’s been considerable debate in the literature, particularly initiated by Jim Darnell: he spoke in past research about the “fallacy of RNA epigenetics”, largely relating to the terminology of the word epigenetics in respect of RNA. The people who created that term did in fact concur, and they say they no longer even use the term, because they recognise it has a problem.
The word “epigenetics” implies inheritance information, from germ cells to progeny. That’s definitely not what RNA modifications do. Our lab introduced the term “epitranscriptomics” in our mapping paper in 2012. The term means simply “on top of the transcriptome” as methyl modifications typically sit on top of the transcriptome.
While the nomenclature is debated most frequently and prominently, it’s the least important of the issues in epitranscriptomics: there are three larger issues which have a real impact on where the field is going from here.
FLG: What Are the Other Controversies Surrounding the Epitranscriptomics Field?
SJ: First, what constitutes the epitranscriptomic code? Some researchers have argued that the fate of mRNA can be determined by many different types of modifications, not just m6A. There is debate surrounding whether these modifications actually exist in mRNA, which sites are actually modified, and if any mRNA modification other than m6A is influential.
There is also discussion on whether or not m6A is dynamic. A lot of the early studies said that the pattern of m6A throughout cellular mRNA changes in different cancers or disease conditions. There was a very important recent paper by Chris Mason published in BioxRv, which argues most of these studies were flawed by not using replicates. He argues that much of the proposed variation in m6A is just statistical noise. A recent paper from Schragi Schwartz published in Cell supports this idea using a very quantitative assay for m6A: the researchers argued that there’s very little or no variation in m6A, and what’s dynamic is the ability of the cell to detect m6A: the m6A readers are dynamic, not the writing of m6A, e.g. where it’s installed.
The last controversy is whether or not m6A erasers actually exist, in particular whether the FTO protein in an m6A eraser. One lab in the m6A field announced that FTO demethylates m6a, raising the possibility that mRNAs can be methylated to m6A, and then converted back to adenosine. However, this theory seems unable to be validated by outside groups, and my lab and others have argued FTO doesn’t even work on mRNA, but instead on small nuclear RNAs. The Schwartz paper published last week in Cell tried to address this controversy by using their new highly quantitative m6A stoichiometry technique called MAZTER-Seq. They mapped all m6A changes that occur in FTO knockouts, and they found no changes were occurring.
The data isn’t looking favourable for the idea that FTO has any biological affects on m6A. We have shown that FTO actually acts on small nuclear RNAs and demethylates a related, but different nucleotide modification called m6Am. In the case of m6A, my lab thinks that m6A erasing is not a major regulatory mechanism. Most of the dynamics seem to relate to reading m6A. But there may be unusual circumstances where writing or erasing m6A is dynamic.
FLG: How Important is Epitranscriptomics for Drug Treatment and Discovery?
SJ: A lot of the interest in drug development emerged with three studies, one published by my lab in collaboration with leukaemia expert Michael Kharas, one by Tony Kouzarides’ lab in the UK and another by Jianjun Chen’s lab at City of Hope, showing leukaemia cells appeared to have elevated levels of m6A writers. When the methyltransferase was inhibited, the leukaemia cells started to differentiate and look like normal cells. Basically, achieving what everyone wants to achieve in leukaemia, which is causing these cells to differentiate.
Because of these studies, multiple companies are starting to focus on how to target the m6A pathway and potentially other epitranscriptomics modifications.
However, if they don’t know the epitranscriptomic code, then they may go after the wrong modifications, especially if they’re going after other, less validated modifications than m6A. But definitely there’s been a lot of interest – some of the larger pharma have programs related to targeting RNA modifications, in particular m6A.
At the end of the day, targeting gene expression is one of the fundamental concepts behind pharma, and RNA modifications are a newly understood way that cells control their gene expression. If you can tell where RNA modifications drive disease phenotypes, then the epitranscriptome may be a fruitful target for drug development.