MBD-Seq of human primary colorectal tumours to explore abnormal CGI methylation. EGA study.
ABSTRACT: A common hallmark of neoplastic cells is abnormal acquisition of CpG island (CGI) methylation and the silencing of associated gene promoters. To determine the extent and distribution of CGI methylation in colorectal cancer, we solexa sequenced MBD affinity purified DNA prepared from five colorectal tumours and matched colon mucosa from the same individuals.
Project description:Tissue-specific methylation patterns suggest a role for CpG island (CGI) methylation in differentiation and cell-type-specific gene regulation. We have applied CXXC affinity purification (CAP), MBD affinity purification (MAP) and chromatin immunoprecipitation (ChIP) in combination with solexa sequencing to investigate the extent and functional significance of CGI methylation in sperm, blood, cerebellum and ES cells for both human and mouse.
Project description:Many human characteristics, including susceptibility to disease, are determined genetically. An unexplored alternative to such genetic determination concerns epigenetic mechanisms such as DNA methylation. CpG islands (CGIs) are generally constitutively hypomethylated, however there are circumstances in which they become heavily methylated and, when coincident with a gene promoter, this invariably causes transcriptional silencing. CGI methylation occurs in normal tissues during processes such as X-inactivation, but abnormal patterns of methylation have also been implicated in disease. The vast majority of evidence relates to cancer, where silencing of multiple genes in this way appears to be a causal contributor to the cancer state. To address the role and extent of CGI methylation in ‘normal’ and diseased cells we applied MBD-affinity purification in conjunction with next generation sequencing in a panel of human brain autopsy samples. These samples represent a panel of individuals as well as specific brain regions and neurological pathologies.
Project description:A common hallmark of neoplastic cells is abnormal acquisition of CpG island (CGI) methylation and the silencing of associated gene promoters. To determine the extent and distribution of CGI methylation in colorectal cancer, we solexa sequenced MBD affinity purified DNA prepared from five colorectal tumours and matched colon mucosa from the same individuals.
Project description:Aim: We aim to compare current (MeDIP-seq), new (Illumina Infinium 450K BeadChip) and future (PacBio) methods for whole genome DNA methylation analysis. As the interest in determination of disease methylation profiles increases, the scope, advantages and limitations of these methods requires assessment. There are key questions to answer and specific challenges to overcome. For example, how much detail/resolution is sufficient to identify regions of differential methylation and regions of biological/medical significance within a sample? How much coverage of the genome is required for accurate methylation analysis? Is it important to confirm which regions of the genome are unmethylated in addition to focusing on those that are methylated? Loss of methylation may be of equal importance within the cell since this may also contribute to disease pathogenesis. A multi-method (affinity enrichment/bisulphite-conversion based/direct sequencing of methyl-cytosine) and technology platform (Illumina HiSeq/PacBio/Illumina Infinium BeadChip) comparison will enable us to determine the strengths and weakness of each method. We propose to compare four methods using two DNA samples from the Coriell Institute for Cell Repository to assess both current and future capabilities for whole genome methylation analysis in parallel: A) MeDIP-seq using Illumina HiSeq B) Illumina Infinium HumanMethylation 450K BeadChip and C) whole genome methylation sequencing using PacBio. Existing single molecule deep bisulphite sequencing data generated previously from these same samples at the WTSI for targeted regions (30-40 genes) on the human X chromosome will be used to assess performance of each method. The methods selected for this study will generate data covering a range of resolutions from a whole genome scan to array (target defined) resolution and up to single base pair, single molecule resolution; the highest level of detail possible with methods currently available.Samples: DNA from sibling pair GM01240 (female) and GM01240 (male).Requirements: Both samples will be analysed using;A.MeDIP-seq using Illumina HiSeq (one HiSeq lane, 75bp paired end, per sample) B.Illumina Infinium HumanMethylation 450K BeadChipWe are expecting a potentially unnecessary high coverage using one HiSeq lane per sample. However, for the MeDIP procedure we do not have a multiplexing procedure in place. Our requirements for PacBio sequencing have been discussed with and will be supported by the Sequencing Technology Development group.
Project description:We propose to definitively characterise the somatic genetics of ER+ve, HER2-ve breast cancer through generation of comprehensive catalogues of somatic mutations in breast cancer cases by high coverage genome sequencing coupled with integrated transcriptomic and methylation analyses.
Project description:Tissue-specific methylation patterns suggest a role for CpG island (CGI) methylation in differentiation and cell-type-specific gene regulation. We have applied CXXC affinity purification (CAP), MBD affinity purification (MAP) and chromatin immunoprecipitation (ChIP) in combination with solexa sequencing to investigate the extent and functional significance of CGI methylation in mouse sperm, blood, cerebellum and ES cells.
Project description:This dataset contains whole-genome MBD (methylbinding domain) sequencing results from cortical neuronal cultures and serves as the basis for characterization of DNA methylation profiles from neuronal systems. This experiment contains three sequencing datasets from 2 biological samples. Two datasets originate from samples that underwent MBD-capture prior to whole-genome sequencing. A third dataset contains non-MBD-captured genomic DNA as a control.
Project description:In order to gain insight into DNA methylation readout, we have established a controlled strategy for profiling genomic targeting of chromatin-interacting factors in vivo. With this approach we determined binding preferences for the methyl-CpG binding domain (MBD) family of proteins, including disease relevant mutants, deletions and isoforms. In vivo binding of MBD proteins occurs as a linear function of local methylation density, and is dependent on functional MBD domain – methyl-CpG interactions. This directs specificity of MBD proteins to methylated, CpG dense and inactive regulatory regions. In contrast, binding to unmethylated sites is MBD protein specific and mediated via alternative domains or protein-protein interactions. The latter is observed for NuRD complex-mediated MBD2 tethering to a subset of unmethylated, tissue-specific regulatory regions, similar to MBD3. These functional binding maps reveal methylation-dependent and -independent binding modes determined by distinct protein domains and revise current models of DNA methylation readout through MBD proteins. Comparative binding analysis for the MBD family of proteins utilizing recombinase-assisted mapping of biotin-tagged proteins (RAMBiO). This set contains maps for 29 samples including wild type MBD proteins, mutants and domain variants in mouse ES cells, derived neurons (NP and TN) and Dnmt1/3a/3b triple-KO ES cells (TKO).
Project description:We report the application of single molecule-based sequencing technology in combination with CXXC affinity purifcation (CAP-seq), MBD affinity purification (MAP-seq) and chromatin immunoprecipitation (ChIP-seq) to generate reciprocal methylation and chromatin modifcation maps in human and mouse. We find that contrary to sequence based prediction methods that humans and mice possess highly equivalent compliments of CpG islands (CGIs). The majority of these CGIs are positive for the active histone modification; H3K4me3 in embryonic stem cells (ES cells) the magnitude of which is correlated with the local density of non-methylated CpG. Approximately half of the human and mouse CGIs are distal to annotated gene promoters, yet more than 40% identify unanticipated transcription start sites as defined by RNA polymerase occupancy and published RNA mapping data. These orphans CGIs preferentially acquire DNA methylation in somatic cells, and this corresponds with a loss of H3K4me3 and RNA polymerase II at these sites. Conversely abnormal CGI methylation found in colorectal tumours showed a distinct distribution relative to that found in normal somatic tissues displaying preferential association with loci marked by H3K27me3 in human ES cells. This study provides a comprehensive functional assessment of CGIs in normal and diseased tissues. Examination of CGI methylation status in human and mouse primary tissues.