Tracking the evolution of cancer methylomes

Single-molecule micromanipulation studies of methylated DNA

Cytosine methylated at the five-carbon position is the most widely studied reversible DNA modification. Prior findings indicate that methylation can alter mechanical properties. However, those findings were qualitative and sometimes contradictory, leaving many aspects unclear. By applying single-molecule magnetic force spectroscopy techniques allowing for direct manipulation and dynamic observation of DNA mechanics and mechanically driven strand separation, we investigated how CpG and non-CpG cytosine methylation affects DNA micromechanical properties. We quantitatively characterized DNA stiffness using persistence length measurements from force-extension curves in the nanoscale length regime and demonstrated that cytosine methylation results in longer contour length and increased DNA flexibility (i.e., decreased persistence length). In addition, we observed the preferential formation of plectonemes over unwound single-stranded "bubbles" of DNA under physiologically relevant stretching forces and supercoiling densities. The flexibility and high structural stability of methylated DNA is likely to have significant consequences on the recruitment of proteins recognizing cytosine methylation and DNA packaging.

Solid phase nucleic acid extraction technique in a microfluidic chip using a novel non-chaotropic agent: dimethyl adipimidate

Here, we present a silicon microfluidic system for the purification and extraction of nucleic acids from human body fluid samples utilizing a dimethyl adipimidate (DMA)-based solid-phase extraction method. We propose DMA, which has been used as an amino-reactive cross-linking agent within cells and proteins, as a non-chaotropic reagent for the capture of nucleic acids to overcome the limitations of existing chaotropic and non-chaotropic techniques such as low binding efficiency, PCR inhibition and so on. DMA contains bi-functional imidoesters that form reversible cross-linking structures with DNA therefore providing a high surface-area to volume ratio for capturing DNA without structurally modifying microfluidic channels. In this work, we have first demonstrated highly efficient capture and purification of genomic DNA (T24 cell line) with DMA using a label-free silicon microring resonator sensor device. In addition, we observed the improvement of the DNA amplification efficiency by using the proposed technique for both the genetic (HRAS) and epigenetic (RARβ) analysis of DNA biomarkers. Particularly, we confirmed that the DMA-based solid-phase extraction technique can be applied for the extraction of genomic DNA with higher purity (p < 0.001) using human body fluids (blood and urine) in silicon microfluidic devices compared to other chaotropic methods. Therefore, the proposed technique would be able to harmonize with a micro-total analysis system platform for the analysis of genetic and epigenetic DNA biomarkers related to human diseases in the field of point-of-care (POC) diagnostic applications.

NGSmethDB: an updated genome resource for high quality, single-cytosine resolution methylomes

The updated release of ‘NGSmethDB’ (http://bioinfo2.ugr.es/NGSmethDB) is a repository for single-base whole-genome methylome maps for the best-assembled eukaryotic genomes. Short-read data sets from NGS bisulfite-sequencing projects of cell lines, fresh and pathological tissues are first pre-processed and aligned to the corresponding reference genome, and then the cytosine methylation levels are profiled. One major improvement is the application of a unique bioinformatics protocol to all data sets, thereby assuring the comparability of all values with each other. We implemented stringent quality controls to minimize important error sources, such as sequencing errors, bisulfite failures, clonal reads or single nucleotide variants (SNVs). This leads to reliable and high-quality methylomes, all obtained under uniform settings. Another significant improvement is the detection in parallel of SNVs, which might be crucial for many downstream analyses (e.g. SNVs and differential-methylation relationships). A next-generation methylation browser allows fast and smooth scrolling and zooming, thus speeding data download/upload, at the same time requiring fewer server resources. Several data mining tools allow the comparison/retrieval of methylation levels in different tissues or genome regions. NGSmethDB methylomes are also available as native tracks through a UCSC hub, which allows comparison with a wide range of third-party annotations, in particular phenotype or disease annotations.

CpGislandEVO: a database and genome browser for comparative evolutionary genomics of CpG islands

Hypomethylated, CpG-rich DNA segments (CpG islands, CGIs) are epigenome markers involved in key biological processes. Aberrant methylation is implicated in the appearance of several disorders as cancer, immunodeficiency, or centromere instability. Furthermore, methylation differences at promoter regions between human and chimpanzee strongly associate with genes involved in neurological/psychological disorders and cancers. Therefore, the evolutionary comparative analyses of CGIs can provide insights on the functional role of these epigenome markers in both health and disease. Given the lack of specific tools, we developed CpGislandEVO. Briefly, we first compile a database of statistically significant CGIs for the best assembled mammalian genome sequences available to date. Second, by means of a coupled browser front-end, we focus on the CGIs overlapping orthologous genes extracted from OrthoDB, thus ensuring the comparison between CGIs located on truly homologous genome segments. This allows comparing the main compositional features between homologous CGIs. Finally, to facilitate nucleotide comparisons, we lifted genome coordinates between assemblies from different species, which enables the analysis of sequence divergence by direct count of nucleotide substitutions and indels occurring between homologous CGIs. The resulting CpGislandEVO database, linking together CGIs and single-cytosine DNA methylation data from several mammalian species, is freely available at our website.