UCL Cancer Institute

High-throughput DNA methylation pipeline

Project Leaders: Dr Lee Butcher & Dr Gareth Wilson

The involvement of DNA methylation in health and disease is well established but not yet fully understood. The aim of this project is to develop novel experimental and computational methods for the analysis of 5-methyl-cytosine (5mC) and 5-hydoxymethyl-cytosine (5hmC).

High-throughput DNA methylation pipeline

Background: The DNA code within an organism is the same no matter which cell type you look at. How then, are specific cellular processes orchestrated by this universal code? The answer, most likely, is through epigenetics – mechanisms that occur around the DNA.

By crude linguistic analogy: if DNA provides the lexicon, epigenetics provides the syntax. In this way, a cell’s capabilities are bound up within the genetic code – the DNA – but the degrees to which these capabilities are realised are governed by epigenetics.

Epigenetic marks punctuate and parse elements of DNA code, effectively regulating the time- and location-specific qualities of gene expression. If epigenetic programming is correct, only the required processes of the cell are activated.

One particularly well-characterised epigenetic modification is DNA methylation, which is indispensible to mammalian genomes. Within mammals, DNA methylation occurs almost exclusively at CpG sites and involves the reversible addition of a methyl group (-CH3) to the carbon-5 position of cytosine nucleotides to form 5-methyl cytosine (5-MeC).

As a heuristic, DNA methylation is inversely related to gene expression. For instance, high DNA methylation (hypermethylation) decreases transcripts, thereby down-regulating or even silencing cellular activity; conversely, the absence of DNA methylation (hypomethylation) may up-regulate or even ‘turn on’ cellular processes.

Many aspects of phenotypic plasticity are governed by DNA methylation, whilst aberrant patterns contribute to disease – most notably, cancer.


Problem: The role of DNA methylation in phenotypic plasticity is well-established but not fully understood. Our understanding is likely to be improved if local methylation events are evaluated within the context of epigenome-wide methylation profiles. This scenario calls for fast, efficient and cost-effective techniques compatible with large cohorts. Here, two challenges present themselves: 1) how do we interrogate thousands of methylation-variable positions (MVPs) scattered throughout the genome? and 2) how do we interrogate MVPs in large cohorts of individuals?

The former burden has been eased with the introduction of a technique called methylated DNA immunoprecipitation or MeDIP^1,2, which involves the use of an antibody against 5-MeC to isolate the methylated fraction of a genome; the latter burden remains because MeDIP is laborious and not amenable to high-throughput applications. To see a two-minute animated overview of how MeDIP works, click below:

Goals: The primary aim of this project is to establish an integrated high-throughput methylation analysis pipeline including automation of the MeDIP assay, LIMS, and application of the BATMAN algorithm³ to quantify absolute methylation levels. To automate MeDIP and assess what MeDIP can detect, we are actively working alongside Diagenode, a world leading biotech who specialise in epigenetic techniques and diagnostics.

MeDIP has already been used to identify thousands of tissue-specific differentially methylated regions (tDMRs) – unique methylation signatures that differentiate one tissue from another⁴. Optimizing MeDIP for high-throughput will allow hundreds to thousands of healthy and diseased samples to be screened, which will identify not only additional tDMRs but diseased-tissue DMRs (dtDMRs). Following high-throughput DNA methylation studies, methylomic data from large clinical case-control cohorts can be integrated with their genomic (SNPs, CNVs etc) and transcriptomic (mRNA, miRNA etc) data to identify clinically significant biomarkers, which can then be used to develop improved diagnostics for disease and aid drug development.

Reference List

1. Weber,M. et al. Chromosome-wide and promoter-specific analyses identify sites of differential DNA methylation in normal and transformed human cells. Nat. Genet. 37, 853-862 (2005).
2. Weber,M. et al. Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome. Nat. Genet. 39, 457-466 (2007).
3. Down,T.A. et al. A Bayesian deconvolution strategy for immunoprecipitation-based DNA methylome analysis. Nat Biotechnol. 2008 Jul;26(7):779-85. PMID: 18612301 (pubmed)
4. Rakyan,V.K. et al. An integrated resource for genome-wide identification and analysis of human tissue-specific differentially methylated regions (tDMRs). Genome Res. 2008 Sep;18(9):1518-29. Epub 2008 Jun 24. PMID: 18577705 (pubmed)

Project Members

Lee M Butcher
Stephan Beck
Gareth A Wilson

Collaborators

Vardhman Rakyan (Institute of Cellular and Molecular Science, Barts [London])